US5963746A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Fully distributed processing memory element
Assignee: International Business Machines Corp
•
Filing: 1995-06-06
•
Granted: 1999-10-05
•
Expires: 2015-06-06
Abstract
Abstract A parallel array processor for massively parallel applications is formed with low power CMOS with DRAM processing while incorporating processing elements on a single chip. Eight processors on a single chip have their own associated processing element, significant memory, and I/O and are interconnected with a hypercube based, but modified, topology. These nodes are then interconnected, either by a hypercube, modified hypercube, or ring, or ring within ring network topology. Conventional microprocessor MMPs consume pins and time going to memory. The new architecture merges processor and memory with multiple PMEs (eight 16 bit processors with 32K and I/O) in DRAM and has no memory access delays and uses all the pins for networking. The chip can be a single node of a fine-grained parallel processor. Each chip will have eight 16 bit processors, each processor providing 5 MIPs performance. I/O has three internal ports and one external port shared by the plural processors on the chip. Significant software flexibility is provided to enable quick implementation of existing programs written in common languages. It is a developable and expandable technology without need to develop new pinouts, new software, or new utilities as chip density increases and new hardware is provided for a chip function. The scalable chip PME has internal and external connections for broadcast and asynchronous SIMD, MIMD and SIMIMD (SIMD/MIMD) with dynamic switching of modes. The chip can be used in systems which employ 32, 64 or 128,000 processors, and can be used for lower, intermediate and higher ranges. Local and global memory functions can all be provided by the chips themselves, and the system can connect to and support other global memories and DASD. The chip can be used as a microprocessor accelerator, in personal computer applications, as a vision or avionics computer system, or as workstation or supercomputer. There is program compatibility for the fully scalable system.
Inventors
- 👤 Thomas Norman Barker
- 👤 Clive Allan Collins
- 👤 Michael Charles Dapp
- 👤 James Warren Dieffenderfer
- 👤 Billy Jack Knowles
- 👤 Donald Michael Lesmeister
- 👤 Richard Ernest Miles
- 👤 Richard Edward Nier
- 👤 Robert Reist Richardson
- 👤 David Bruce Rolfe
- 👤 Vincent John Smoral
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-31 08:05
CPC Classifications
G; G06; G06F; G06F15/00; G06F15/76; G06F15/80; G06F15/8007
Claims Analysis
Total Claims: 158
Claim 1
Independent Claim
248 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory; each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate; each of said processor memory elements having a single respective external communication path coupled directly to a respective one of said plurality of processing memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, each processor memory element having an independent router function operative in the communication over said internal communication path and said single respective external communication path; and wherein at least two of said processor memory elements on the same common substrate are capable of concurrently communicating over their said single respective external communication path to respective processor memory elements independent of a synchronous time relationship for their respective communications; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 2
Dependent Claim
248 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory; each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate; each of said processor memory elements having a single respective external communication path coupled directly to a respective one of said plurality of processing memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, each processor memory element having an independent router function operative in the communication over said internal communication path and said single respective external communication path; and wherein at least two of said processor memory elements on the same common substrate are capable of concurrently communicating over their said single respective external communication path to respective processor memory elements independent of a synchronous time relationship for their respective communications; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 3
Dependent Claim
46 words
The computer system according to claim 1 wherein said common substrate forms a network node having eight or more of said processor memory elements wherein node communication is provided by fully distributed I/O routers and signal I/O ports at each of the single processor memory elements.
Claim 4
Dependent Claim
15 words
The computer system according to claim 1 wherein said common substrate is a CMOS chip.
Claim 5
Dependent Claim
39 words
The computer system according to claim 1 wherein said common substrate is a chip and the internal and external communication paths plus the functional relationships between the processor memory elements and their communication paths are provided on said chip.
Claim 6
Dependent Claim
40 words
The computer system according to claim 1 wherein the processor memory element has internal data flows using high speed hard registers to feed distributed ALU and I/O router registers and necessary logic within said processor memory element for all operations.
Claim 7
Dependent Claim
51 words
The computer system according to claim 1 wherein a node which includes a plurality of said processor memory elements is connected to other nodes in a cluster network with data routing characterized by a task sequencing function, distributed between hardware and software, with software controlling most of the task sequencing function.
Claim 8
Dependent Claim
62 words
The computer system according to claim 1 wherein a node which includes a plurality of said processor memory elements is connected to other nodes in a cluster network with data routing distributed between hardware and software in each of said processor memory elements, with hardware provided for performing inner loop transfers and minimizing overhead for performing outer loop transfers of the node.
Claim 9
Dependent Claim
71 words
The computer system according to claim 1 wherein a node of a processor array includes a plurality of said processor memory elements, said dedicated local memory including 32K 16-bit words addressable by and coupled to said processor, the processor comprising a 16-bit processor, each said processor memory element having an I/O router and a signal I/O port, thereby providing for a node with fully distributed I/O routers and signal I/O ports.
Claim 10
Dependent Claim
63 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster with parallel communication paths between the processor memory elements along said internal and external communication paths providing a processing array, and wherein the system includes means for supporting distinct user levels ranging from completely unaware to fully involved in detailed tasking.
Claim 11
Dependent Claim
67 words
The computer system according to claim 1 organized as an n-dimensional network of chip processor memory elements having a predetermined number of pins per chip wherein I/O is distributed over as many pins as available while memory associated with each processor memory element is accessible to other processor memory elements providing scalable and large memory by replication of nodes of the system provided with parallel I/O bandwidths.
Claim 12
Dependent Claim
49 words
The computer system according to claim 1 further comprising means for enabling distributed fault tolerance by an n-redundancy mode, where the data is given a redundancy number based on its importance and replicated, a number of times according to the redundancy number, within an array of processor memory elements.
Claim 13
Dependent Claim
88 words
The computer system according to claim 1, further comprising a cluster controller and an application processor interface, said cluster controller controlling a plurality of nodes, each node including at least one of said processor memory elements, the nodes forming a processing array, said application processor interface coupling said processing array within a computer system, and wherein, among said application processor interface, said cluster controller, and said processor memory elements, control tasks are distributed to multiple levels and highly repetitive tasks are distributed to low levels of the system.
Claim 14
Dependent Claim
53 words
The computer system according to claim 1 further comprising within a processor memory element: a native instruction set, means for providing an expandable multiply function, a programmable router for routing information alternatively left/right, NEWS matrix, NEWS/up-down, hypercube, and wherein said programmable router employs a hardwired distributed router provided by each processor memory element.
Claim 15
Dependent Claim
46 words
The computer system according to claim 1 wherein said processor memory elements are replicated on said common substrate as a single chip design that is utilized as computational unit used as a main building block for system ranging from PC accelerators to super computer size machines.
Claim 16
Dependent Claim
43 words
The computer system according to claim 1 wherein each of a plurality of said common substrate of processor memory elements has 2n processors, memory and routers within a single chip and internal and external communication paths, thereby minimizing delays due to chip crossings.
Claim 17
Dependent Claim
59 words
The computer system according to claim 1 wherein each of a plurality of said common substrate of processor memory elements has a processor, memory and routers within a single chip and internal and external communication paths, thereby minimizing delays due to chip crossings, each processor memory element having means for fixed and floating point processing, routing and I/O control.
Claim 18
Dependent Claim
21 words
The computer system according to claim 1 wherein each said processor memory element provides a program interrupt capability for the system.
Claim 19
Dependent Claim
27 words
The computer system according to claim 1 wherein each said processor memory element includes 64K bytes of dedicated local memory, and provides a performance of 2.0 MIPS.
Claim 20
Dependent Claim
59 words
The computer system according to claim 1 wherein said plurality of processor memory elements on said common substrate forms a packaged single chip of eight of said processors each coupled to said dedicated local memory and each having associated communication logic which enables a chip to be defined as a node in an array characterized by a node architecture.
Claim 21
Dependent Claim
43 words
The computer system according to claim 1 wherein each processor memory element comprises a 16-bit processor memory element with full 16-bit internal control and data flow for communication between the processor memory elements within a network wherein said common substrate forms a node.
Claim 22
Dependent Claim
55 words
The computer system according to claim 1 wherein the common substrate is a chip and eight of said processor memory elements are provided on said chip, said chip providing a node of system composed of a plurality of nodes, and wherein processor memory elements within the node are connected in a binary hypercube communication network.
Claim 23
Dependent Claim
46 words
The computer system according to claim 1 wherein the common substrate is a chip and eight processor memory elements are provided on said chip, and communication between processor memory elements is controlled by in and out registers under control of each of the processor memory element.
Claim 24
Dependent Claim
88 words
The computer system according to claim 1 wherein the common substrate is a chip and eight processor memory elements are provided on said chip, said chip providing a node of a node array composed of a plurality of nodes, and wherein processor memory elements within the node are connected in a binary hypercube communication network, and each node has eight processor memory elements on said chip, and communication between processor memory elements is controlled by in and out registers under control of each of the processor memory elements.
Claim 25
Dependent Claim
60 words
The computer system according to claim 1 further comprising a system controller, and a broadcast control interface coupled to each processor memory element, said broadcast control interface enabling the system controller and the processor memory elements to interrogate one of said processor memory elements for status information within the processor memory element even during operation of said processor memory element.
Claim 26
Dependent Claim
50 words
The computer system according to claim 1 further comprising a system controller, and a broadcast control interface coupled to each of said processor memory elements, said broadcast control interface providing means for providing inter-processor busing protocols, said system controller functioning as an array controller for nodes of processor memory elements.
Claim 27
Dependent Claim
49 words
The computer system according to claim 1 further comprising a system controller, and a broadcast control interface coupled to each of said processor memory elements, said broadcast control interface providing means for providing inter-processor memory element busing protocols, said system functioning as an array accelerator for a conventional processor.
Claim 28
Dependent Claim
40 words
The computer system according to claim 1 further comprising means for enabling clustering of a set of the processor memory elements within a node, the node comprising a plurality of processor memory elements, for sharing common broadcast and control interfaces.
Claim 29
Dependent Claim
36 words
The computer system according to claim 1 further comprising within a processor memory element: means for handling interrupts of differing levels; multiple register sets for each interrupt level providing means for saving status when taking interrupts.
Claim 30
Dependent Claim
38 words
The computer system according to claim 1 further comprising within each processor memory element: storage register elements coupled to the processor and mapped to the dedicated local memory for enabling hardware expansion of memory without a software impact.
Claim 31
Dependent Claim
35 words
The computer system according to claim 1 wherein the communication paths are wrapped to connect nodes of the computer system as cubes, each of the nodes including at least one of said processor memory elements.
Claim 32
Dependent Claim
28 words
The computer system according to claim 1 wherein said processor and said internal and external communication paths include means for providing four interrupts levels and two communication modes.
Claim 33
Dependent Claim
18 words
The computer system according to claim 31 wherein one of the two communication modes permits direct memory addressing.
Claim 34
Dependent Claim
28 words
The computer system according to claim 1 wherein communication modes between processor memory elements are provided for handling contention with busy processor memory elements in a data stream.
Claim 35
Dependent Claim
45 words
The computer system according to claim 1 wherein a node includes at least one of said processor memory elements, and connections between the nodes of the computer system provides NEWS connections, 2-d torus connections, and 4-D wrapped torus connections among nodes of the computer system.
Claim 36
Dependent Claim
66 words
The computer system according to claim 1 organized as a massively parallel machine, wherein a node includes at least one of said processor memory elements, with nodes interconnected as a n dimensional hypercube network with parallel communication paths between the processor memory elements along said internal and external communication paths providing N-dimensional networking among the nodes, the interconnection among the nodes supporting a 4 dimensional hypercube.
Claim 37
Dependent Claim
82 words
The computer system according to claim 1 organized as a massively parallel machine, wherein a node includes at least one of said processor memory elements, with nodes interconnected as a n dimensional hypercube network with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and further comprising a set of modules which serve as a modular array controller for controlling the processing array such that varying size systems can be configured by replication.
Claim 38
Dependent Claim
74 words
The computer system according to claim 1 organized as a massively parallel machine, wherein a node includes at least one of said processor memory elements, with nodes interconnected as a n dimensional network cluster with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and further comprising an array controller for controlling the processing array such that varying size systems can be configured by replication.
Claim 39
Dependent Claim
97 words
The computer system according to claim 1 organized as a massively parallel machine, wherein a node includes at least one of said processor memory elements, with nodes interconnected as a n dimensional network cluster with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein each node of a processing array has eight processor memory elements on a chip which are interconnected as part of a system network, and node to node control includes a broadcast bus and tag lines to provide a system interconnection among nodes.
Claim 40
Dependent Claim
63 words
The computer system according to claim 1 organized as a massively parallel machine, wherein a node includes at least one of said processor memory elements, with nodes interconnected as a n dimensional network cluster with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein each node interconnection is an eight element hypercube.
Claim 41
Dependent Claim
17 words
The computer system according to claim 1 wherein each processor memory element provides for floating point calculation.
Claim 42
Dependent Claim
26 words
The computer system according to claim 1 wherein each processor memory element provides for floating point calculation by emulating floating point on a processor memory element.
Claim 43
Dependent Claim
27 words
The computer system according to claim 1 wherein each processor memory element provides for floating point calculation by a hardware floating point accelerator section provided by logic.
Claim 44
Dependent Claim
43 words
The computer system according to claim 1 organized as an n-dimensional network of processor memory elements which selectively performs MIMD operations and wherein an array of nodes is provided with directed inter-processor memory element communication for performing sparse to dense data structure transformations.
Claim 45
Dependent Claim
56 words
The computer system according to claim 1 organized as an n-dimensional network of processor memory elements which selectively performs MIMD operations and wherein an array of nodes is provided with means for directed inter-processor memory element communication for preparing data for pipelined very high speed arithmetic and for mapping results back to a sparse data structure.
Claim 46
Dependent Claim
41 words
The computer system according to claim 1 organized as an n-dimensional network of processor memory elements which selectively performs MIMD operations and wherein an array of nodes is provided with means for data packing/unpacking for high speed pipelined floating point operations.
Claim 47
Dependent Claim
54 words
The computer system according to claim 1 organized as an n-dimensional network of the processor memory elements which selectively performs MIMD operations and wherein an array of nodes, each of the nodes including at least one of the processor memory elements, is provided with means for performing scatter/gather operations in parallel with arithmetic operations.
Claim 48
Dependent Claim
52 words
The computer system according to claim 1 organized as an n-dimensional network of processor memory elements which selectively performs MIMD operations and wherein an array of nodes, each of the nodes including at least one of the processor memory elements, is provided with means for performing programmable support for sparse matrix encoding.
Claim 49
Dependent Claim
24 words
The computer system according to claim 1 organized as an n-dimensional network of processor memory elements having memory resident registers and floating point capability.
Claim 50
Dependent Claim
24 words
The computer system according to claim 1 organized as an n-dimensional network of chip processor memory elements having chip I/O pins utilized for control.
Claim 51
Dependent Claim
60 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein said system has multiple clusters.
Claim 52
Dependent Claim
69 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein said system has quarter cluster wiring in four dimensions characterized by X,Y,W, and Z.
Claim 53
Dependent Claim
92 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein said system has a cluster with internal wiring in the X and Y dimensions, and external wiring in the W and Z dimensions, X, Y, W, and Z characterizing four independent dimensions of said massively parallel machine.
Claim 54
Dependent Claim
68 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein said system has cluster to cluster wiring in the W and Z dimensions.
Claim 55
Dependent Claim
86 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein said system is provided with an array director having overall control of the network cluster, said array director including a cluster controller for each cluster comprising an array of the nodes.
Claim 56
Dependent Claim
92 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein said system is provided with a plurality of clusters and an array director having overall control of the clusters providing a cluster controller for an array of nodes, and a cluster synchronizer for an array of clusters.
Claim 57
Dependent Claim
118 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein said system is provided with a plurality of clusters, each cluster having edges, and an array director having overall control of the clusters providing a cluster controller for an array of nodes, and a cluster synchronizer for an array of clusters, and an I/O zipper along the edges of selected clusters for selectively inputting data to and outputting data from the network cluster.
Claim 58
Dependent Claim
120 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein said system is provided with a plurality of clusters, each cluster having edges, and an array director having overall control of the clusters providing a cluster controller for an 8×8 array of nodes, and a cluster synchronizer for an 8×8 array of clusters, and an I/O zipper along the edges of selected clusters for selectively inputting data to and outputting data from the network cluster.
Claim 59
Dependent Claim
97 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network, wherein a node includes at least one of said processor memory elements, with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein said system has rings for communication which can be broken to provide an effective interface to external I/O bus paths, and interfaces to attach external storage or additional display units to a subset of processor memory elements of the processing array of nodes.
Claim 60
Dependent Claim
72 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein any set of inter-processor memory element communication paths can be logically broken to selectively provide external I/O.
Claim 61
Dependent Claim
85 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein any set of inter-processor memory element communication paths can be broken to add an external I/O interface with all processor memory elements of a node interfacing to the external interface.
Claim 62
Dependent Claim
99 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein the processor memory elements of an array have a means for providing a transparent mode that is utilized when routing data between processor memory elements within a chip set of processor memory elements, thereby reducing the effective network diameter of a network of nodes.
Claim 63
Dependent Claim
77 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein simulation and users are supported with variable resolution simulator software tools and support software visualization tools for load balancing and performance evaluation.
Claim 64
Dependent Claim
87 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, said computer system further comprising an application program interface for coupling said processing array within a computer system, with a processor and software within the application processor interface enabling system utilization to be controlled.
Claim 65
Dependent Claim
102 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, said computer system further comprising an application program interface for coupling said processing array within a computer system, with a processor and software within the application processor interface enabling system utilization to be controlled in either a standalone mode or as an attached processor to one or more hosts.
Claim 66
Dependent Claim
90 words
The computer system according to claim 1 organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster with parallel communication paths between processor memory elements along said internal and external communication paths providing a processing array, said computer system further comprising an application program interface for coupling said processing array within a computer system, with a processor and software within the application processor interface enabling system utilization to be controlled when coupled to one or more hosts by either a tightly or loosely coupled attachment.
Claim 67
Dependent Claim
72 words
The computer system according to claim 1 wherein a node element of multiple processor memory elements of a parallel array processor incorporates up to 32,768 identical processor memory elements in an array cluster, said array processor further including an array director, said array director comprising an application processor interface, a cluster synchronizer and cluster controller, the cluster controller coupled to the array cluster, and said cluster synchronizer coupled to the cluster controller.
Claim 68
Dependent Claim
42 words
The computer system according to claim 1 wherein a node element of multiple processor memory elements of a parallel array processor incorporates up to 64 array clusters as the basic building block of a parallel array computer system having an array director.
Claim 69
Dependent Claim
61 words
The computer system according to claim 1 wherein a node element, comprising a plurality of said processor memory elements of a parallel array processor incorporates array clusters as a basic building block of a parallel array computer system having an array director for controlling clusters, and an application processor interface that supports a test/debug unit for design, debug, and monitoring functions.
Claim 70
Dependent Claim
67 words
The computer system according to claim 1 wherein a node element comprising a plurality of said processor memory elements, of a parallel array processor forms part of a cluster, each node having a set of multiple processor memory elements, and wherein a complete set of processor memory element may be divided into 1 to many distinct subsets under control of an array director to interleave between subsets.
Claim 71
Dependent Claim
79 words
The computer system according to claim 1 wherein the system has clusters formed from arrays of nodes, each node having a set of multiple processor memory elements, and wherein a complete set of processor memory element may be divided into 1 to many distinct subsets under control of an array director to interleave steps between subsets, the step sequence of the interleave process and the amount of control exercised over each subset being program controlled by said array director.
Claim 72
Dependent Claim
60 words
The computer system according to claim 1 wherein a complete set of processor memory elements may be divided into 1 to many distinct subsets under control of an array director which interleaves between subsets to operate distinct subsets of the array in one mode while other sets operate in tightly synchronized SIMD mode under the control of the array director.
Claim 73
Dependent Claim
44 words
The computer system according to claim 1 wherein processor memory elements have 16-bit wide dataflow by 16-bit memory, I/O ports and I/O switching paths, and control logic to permit any processor memory element to fetch, decode and execute a 16 bit reduced instruction set.
Claim 74
Dependent Claim
72 words
The computer system according to claim 1 wherein processor memory elements have I/O ports and I/O switching paths, and control logic to permit any processor memory element to fetch, decode and execute a reduced instruction set, and further logic enables a processor memory element to perform as both a processing unit and data router with specialized control means for rapid task switching and retention of program state information at interrupt execution levels.
Claim 75
Dependent Claim
47 words
The computer system according to claim 1, further comprising a clock, and wherein a processor memory element in SIMD mode receives decoded instruction information in synchronization with a given cycle of the clock and executes that data in a clock cycle immediately subsequent to the given cycle.
Claim 76
Dependent Claim
35 words
The computer system according to claim 1 wherein a processor memory element in MIMD mode maintains a program counter PC address and uses that as an address within its own memory to fetch an instruction.
Claim 77
Dependent Claim
31 words
The computer system according to claim 1 wherein a processor memory element has three input ports, and three output ports for on-chip communication plus an I/O port for off chip communications.
Claim 78
Dependent Claim
24 words
The computer system according to claim 1 wherein a processor memory element has an I/O port for off substrate byte wide half duplex communication.
Claim 79
Dependent Claim
61 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has an I/O port for off substrate byte wide communication, and has input ports that are connected such that data may be routed from input to memory, or from an input address register to an output register via a direct parallel data path.
Claim 80
Dependent Claim
77 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has broadcast and control interface provided on each substrate, thereby providing a parallel input interface and also providing two serial port interfaces, the serial port interfaces providing each processor memory element with the capability to output a limited amount of status information for signaling an array director when the processor memory element has data that needs to be read.
Claim 81
Dependent Claim
78 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has broadcast and control interface provided on each substrate, thereby providing a parallel input interface and also providing two serial port interfaces, the serial port interfaces providing each processor memory element with the capability to output activity status information such that test and monitor displays can illustrate the status of the computer system of which a node forms a part.
Claim 82
Dependent Claim
75 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has broadcast and control interface provided on each substrate, thereby providing a parallel input interface and also providing two serial port interfaces, the serial port interfaces providing a serial port to permit an array director to input control data so as to direct the port to particular processor memory element and node internal registers and access points.
Claim 83
Dependent Claim
39 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has multiple processor memory elements and each processing element is enabled to process data lengths which are multiples of 16 bits.
Claim 84
Dependent Claim
72 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has broadcast and control interface provided on each substrate, and wherein the broadcast and control interface is coupled to the broadcast and control interfaces of each of a set of nodes and to a node controller via a serial scan loop for acquiring information on individual nodes or processor memory elements within a node.
Claim 85
Dependent Claim
81 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has broadcast and control interface provided on each substrate, and wherein the broadcast and control interface is coupled to a node controller coupled through a set of nodes with a serial scan loop for acquiring information on individual nodes or processor memory elements within a node, with loops initially interfacing to the broadcast and control interface which bridges to individual processor memory elements.
Claim 86
Dependent Claim
62 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has multiple processor memory elements with single instructions which selectively performs adds/subtracts/compares/moves on operands of length 1 to 8 words where a length is a part of the instruction, enabling a programmer to view an operand length as 16 to 128 bits in length.
Claim 87
Dependent Claim
87 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has multiple processor memory elements with single instructions which selectively performs adds/subtracts/compares/moves on operands of length 1 to 8 words where a length is a part of the instruction, enabling a programmer to view an operand length as 16 to 128 bits in length, an by using combinations of these instructions, operands of any length can be manipulated to add two numbers of up to 256 bits in length.
Claim 88
Dependent Claim
78 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has multiple processor memory elements and is configured within the node as a 2 by n array with each processor memory element interconnected with its three neighbors using a set of input/output ports, thus, providing full duplex communication capability between processor memory elements, with each processor memory element having an external input and output port connected to node I/O pins.
Claim 89
Dependent Claim
39 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has external I/O pins for external to node communication with input and output ports connected to share pins for half-duplex communication.
Claim 90
Dependent Claim
39 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has external I/O pins for external to node communication with input and output ports connected to separate pins for full-duplex communication.
Claim 91
Dependent Claim
35 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has eight processor memory elements within the node which are connected in a binary hypercube communication network.
Claim 92
Dependent Claim
74 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has eight processor memory elements within the node which are connected in a binary hypercube communication network, wherein each processor memory element within the node is connected to three other processor memory elements within the node, and wherein communication among processor memory elements is controlled by in and out registers under control of the processing element.
Claim 93
Dependent Claim
58 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has multiple processor memory elements with an inter-processor memory element network that is a 4d torus mechanism used for inter-processor memory element communication whereby any processor memory element can reach any other processor memory elements configured in an array processor.
Claim 94
Dependent Claim
176 words
The computer system according to claim 1 wherein a processor memory element forms part of a node said node comprising: a. eight processor memory elements forming a chip, each of said processor memory elements including a 16-bit processor and 32K words in 64K bytes memory, b. a broadcast and control interface enabling an external controller to operate all or subsets of the processor memory elements and to accumulate processor memory element requests, and having plural interconnection levels whereby: 1) each processor memory element supports four 8-bit wide inter-processor memory element communication paths which connect to three neighboring processor memory elements on the chip and one off chip processor memory element, 2) broadcast-to-processor memory element busing, which makes data or instructions available to processor memory elements, 3) service request lines that permit any processor memory element to send a code to the external controller, and 4) serial service loops permitting the external controller to read processor memory element status information with an interconnection level which extends from the node to all processor memory elements within the node.
Claim 95
Dependent Claim
44 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has multiple processor memory elements on a chip with interconnections to their immediate neighbors with an interconnection pattern that results in a three dimensional cube structure.
Claim 96
Dependent Claim
63 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has multiple processor memory elements on a chip with interconnections to their immediate neighbors with an interconnection pattern that results in a three dimensional cube structure, and wherein each processor memory element within the cube has a dedicated bidirectional byte port to the chip's pins.
Claim 97
Dependent Claim
76 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which forms part of the computer system in which an n by n array of nodes is a cluster, and bridging between + x and - x ports and the + y and - y ports provide the cluster node interconnections, wherein n is a positive integer and wherein x an y represent independent dimensions of said computer system.
Claim 98
Dependent Claim
78 words
The computer system according to claim 1 wherein the processor memory element forms part of a node which forms part of the computer system in which an n by n array of nodes is a cluster, and bridging between the + and - x ports and the + and - y ports provide the cluster node interconnections, and each of the nodes has a plurality of said processor memory elements, each of which manages a single external port.
Claim 99
Dependent Claim
92 words
The computer system according to claim 1 wherein the processor memory element forms part of a node which forms part of the computer system in which an n by n array of nodes is a cluster, and bridging between the + and - x ports and the + and - y ports provide the cluster node interconnections, and each of the nodes has a plurality of said processor memory elements, each of which manages a single external port, and node bridging at the outer edges makes the cluster into a logical torus.
Claim 100
Dependent Claim
70 words
The computer system according to claim 1 wherein the processor memory element forms part of a node which forms part of the computer system in which an n by n array of nodes is a cluster, and bridging between the + and - w ports and + and - z ports provides 4d torus interconnections with each node within a cluster connected to an equivalent node in all adjacent clusters.
Claim 101
Dependent Claim
83 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which forms part of the computer system in which an n by n array of nodes is a cluster, and bridging between the + and - w ports and + and - z ports provides the 4d torus interconnections with each node within a cluster connected to an equivalent node in all adjacent clusters for providing 64 ports between two adjacent clusters of an array processor.
Claim 102
Dependent Claim
43 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has multiple processor memory elements coupled by an optical path and an on-substrate optical driver/receiver for coupling of nodes in an array of 4d toruses.
Claim 103
Dependent Claim
45 words
The computer system according to claim 1 wherein a node processor memory element forms part of a node which has twelve processor memory elements on a chip with an on-chip optical driver/receiver for coupling of nodes in an array of 4d toruses having multi-flop performance.
Claim 104
Dependent Claim
44 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has multiple processor memory elements and is connected to other nodes in a cluster network with I/O programs at dedicated interrupt levels for managing the network.
Claim 105
Dependent Claim
108 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has multiple processor memory elements and is connected to other nodes in a cluster network with I/O programs at dedicated interrupt levels for managing the network, each processor memory element having interrupt registers and dedicating four interrupt levels to receiving data from four neighbors, a buffer provided at each level by loading registers at the level, and having IN and RETURN instruction pairs using a buffer address and transfer count to enable the processor memory element to accept words from an input bus and to store them to the buffer.
Claim 106
Dependent Claim
159 words
The computer system according to claim 1 wherein a processor memory element forms part of a node which has multiple processor memory elements and is connected to other nodes in a cluster network with I/O programs at dedicated interrupt levels for managing the network, each processor memory element having a program counter, interrupt registers and dedicating four interrupt levels to receiving data from four neighbors, a buffer provided at each level by loading registers at the level, and having in and return instruction pairs using a buffer address and transfer count to enable the processor memory element to accept words from an input bus and to store them to the buffer, means enabling a buffer full condition to generate an interrupt and restore the program counter to the instruction after a return to enable I/O programs which do need to test what caused an interrupt to read data, return, and then continue directly into processing the data they read.
Claim 107
Dependent Claim
101 words
The computer system according to claim 1 organized as a massively parallel machine having nodes interconnected as a n dimensional network of clusters with parallel communication paths between said processor memory elements along said internal and external communication paths providing a processing array, each of the nodes including at least one of said processor memory elements, each of said clusters including an array of said processor memory elements, and the external communication paths forming edges of the clusters, and wherein loading or unloading the processing array is done by connecting the edges of at least one cluster to a system bus.
Claim 108
Dependent Claim
96 words
The computer system according to claim 1 wherein each of a plurality of nodes includes at least one of said processor memory elements and data transfers between two of said nodes occurs via a connection directly attached to a zipper interface, said zipper interface including means for selectively providing data transfer into and out of the two nodes, said zipper interface having a number of the selectively enabled connections to a respective number of pairs of the nodes to satisfy application requirements for a given configuration of said processor memory elements, thereby providing different zipper sizes.
Claim 109
Dependent Claim
103 words
The computer system according to claim 1 wherein each of a plurality of nodes includes at least one of said processor memory elements, the nodes forming a node array which allows data transfers into or out of the node array to alternate between two of the nodes directly attached to a zipper interface into the node array, the computer system executing a program, said zipper interface having a width size that is varied depending on a given configuration of said processor memory elements, and depending on the program requirements to roll data and program into and out of the individual processor memory elements.
Claim 110
Dependent Claim
63 words
The computer system according to claim 1 including a zipper connection for a processor array of said processor memory elements, said zipper connection provided as a separate interface into said processor array, said zipper connection being established on one or two edges of a hypercube network by dotting onto a broadcast bus at multiple nodes in the processor array and in multiple directions.
Claim 111
Dependent Claim
107 words
The computer system according to claim 1, wherein a node is formed from at least one of said processor memory elements, and an array processor includes a plurality of nodes, said computer system including an array director having: an application processor interface for coupling said array processor array within a computer system, a cluster synchronizer for managing the operations of an array of the nodes to provide synchronized operation, a cluster controller, coupled to said cluster synchronizer, for controlling input/output operations for an array of the nodes, and a fast I/O for a zipper connection which provides data input to or data output from said array processor.
Claim 112
Dependent Claim
80 words
The computer system according to claim 1, wherein a node is formed from at least one of said processor memory elements, and an array processor includes a plurality of nodes, said computer system including an array controller having: a. an application processor interface for coupling said array processor array within a computer system, b. a cluster controller for controlling an array of nodes, and c. an I/O means for a logically breaking inter-node coupling upon direction of the array controller.
Claim 113
Dependent Claim
95 words
The computer system according to claim 1 wherein a plurality of said processor memory elements form an array processor, and wherein an application processor interface is provided for managing a path to and from an I/O zipper which provides data input to or data output from said array processor, and wherein data received from a host system in attached modes, or from devices in standalone modes is forwarded to the array processor, and wherein prior to initiating this type of operation processor memory elements within an array which will be managing the I/O are initiated.
Claim 114
Dependent Claim
35 words
The computer system according to claim 1, wherein one of the processor memory elements on the substrate addresses the dedicated local memory of another processor memory element of said system, thereby providing global memory addressing.
Claim 115
Dependent Claim
135 words
A computer system comprising: a plurality of scalable nodes, each of said scalable nodes having multiple processor memory elements, each of said processor memory elements having a plurality of communication paths for communication within a node to other processor memory elements within the node, and each processor memory element having a dedicated, single communication path for communication external to said node to another of said scalable nodes of the computer system; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 116
Dependent Claim
135 words
A computer system comprising: a plurality of scalable nodes, each of said scalable nodes having multiple processor memory elements, each of said processor memory elements having a plurality of communication paths for communication within a node to other processor memory elements within the node, and each processor memory element having a dedicated, single communication path for communication external to said node to another of said scalable nodes of the computer system; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 117
Dependent Claim
256 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, each processor memory element having an independent router function operative in the communication over said internal communication path and said dedicated, single external communication path, and wherein each plurality of processor memory elements on the same common substrate has the processor, the memory and the router function provided within a single chip, thereby minimizing delays due to chip crossings, each processor memory element having means for fixed and floating point processing, routing and I/O control; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 118
Dependent Claim
256 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, each processor memory element having an independent router function operative in the communication over said internal communication path and said dedicated, single external communication path, and wherein each plurality of processor memory elements on the same common substrate has the processor, the memory and the router function provided within a single chip, thereby minimizing delays due to chip crossings, each processor memory element having means for fixed and floating point processing, routing and I/O control; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 119
Dependent Claim
201 words
A computer system, comprising, a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and a system controller, and a broadcast control interface, said broadcast control interface providing means for providing inter-processor busing protocols; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 120
Dependent Claim
201 words
A computer system, comprising, a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and a system controller, and a broadcast control interface, said broadcast control interface providing means for providing inter-processor busing protocols; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 121
Dependent Claim
210 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and wherein the communication paths are wrapped to connect nodes of the computer system as cubes, each of the nodes including at least one of said processor memory elements; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 122
Dependent Claim
210 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and wherein the communication paths are wrapped to connect nodes of the computer system as cubes, each of the nodes including at least one of said processor memory elements; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 123
Dependent Claim
203 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and wherein said processor and said internal and external communication paths included means for providing four interrupts levels and two communication modes; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 124
Dependent Claim
203 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and wherein said processor and said internal and external communication paths included means for providing four interrupts levels and two communication modes; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 125
Dependent Claim
254 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, said computer system organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein said system has a cluster with internal wiring in the X and Y dimensions, and external wiring in the W and Z dimensions; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 126
Dependent Claim
254 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, said computer system organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein said system has a cluster with internal wiring in the X and Y dimensions, and external wiring in the W and Z dimensions; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 127
Dependent Claim
289 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, said computer system organized as a massively parallel machine with nodes interconnected as an dimensional network, wherein a node includes at least one of said processor memory elements, with communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein said system has rings for communication which can be broken to provide and effective interface to external I/O bus paths, and interfaces to attach external storage or additional units to a subset of processor memory elements of the processing array of nodes; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 128
Dependent Claim
289 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, said computer system organized as a massively parallel machine with nodes interconnected as an dimensional network, wherein a node includes at least one of said processor memory elements, with communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein said system has rings for communication which can be broken to provide and effective interface to external I/O bus paths, and interfaces to attach external storage or additional units to a subset of processor memory elements of the processing array of nodes; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 129
Dependent Claim
247 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, said computer system organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein any set of inter-processor memory element communication paths can be broken to add an external I/O; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 130
Dependent Claim
247 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, said computer system organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein any set of inter-processor memory element communication paths can be broken to add an external I/O; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 131
Dependent Claim
275 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, the computer system organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein the processor memory elements of an array have a means for providing a transparent mode that is utilized when routing data between processor memory elements within a chip set of processor memory elements for reducing the effective network diameter of a network of nodes; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 132
Dependent Claim
275 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, the computer system organized as a massively parallel machine with nodes interconnected as a n dimensional network cluster, wherein a node includes at least one of said processor memory elements, with communication paths between processor memory elements along said internal and external communication paths providing a processing array, and wherein the processor memory elements of an array have a means for providing a transparent mode that is utilized when routing data between processor memory elements within a chip set of processor memory elements for reducing the effective network diameter of a network of nodes; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 133
Dependent Claim
246 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and wherein processor memory elements have I/O ports and I/O switching paths, and necessary control logic to permit any processor memory element to fetch, decode and execute an instruction, and further logic enables a processor memory element to perform as both a processing unit and data router with specialized control means for rapid task switching and retention of program state information at interrupt execution levels; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 134
Dependent Claim
246 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and wherein processor memory elements have I/O ports and I/O switching paths, and necessary control logic to permit any processor memory element to fetch, decode and execute an instruction, and further logic enables a processor memory element to perform as both a processing unit and data router with specialized control means for rapid task switching and retention of program state information at interrupt execution levels; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 135
Dependent Claim
251 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and wherein, processor memory element forms part of a node which has broadcast and control interface provided on each substrate, thereby providing a parallel input interface and also providing two serial port interfaces, the serial port interfaces providing each processor memory element with the capability to output a limited amount of status information for signaling an array director when the processor memory element has data that needs to be read; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 136
Dependent Claim
251 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and wherein, processor memory element forms part of a node which has broadcast and control interface provided on each substrate, thereby providing a parallel input interface and also providing two serial port interfaces, the serial port interfaces providing each processor memory element with the capability to output a limited amount of status information for signaling an array director when the processor memory element has data that needs to be read; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 137
Dependent Claim
233 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and wherein a processor memory element forms part of a node which has multiple processor memory elements with an inter-processor memory element network that is a 4d torus mechanism used for inter-processor memory element communication whereby any processor memory element can reach any other processor memory elements configured in an array processor; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 138
Dependent Claim
233 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and wherein a processor memory element forms part of a node which has multiple processor memory elements with an inter-processor memory element network that is a 4d torus mechanism used for inter-processor memory element communication whereby any processor memory element can reach any other processor memory elements configured in an array processor; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 139
Dependent Claim
252 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and wherein, processor memory element forms part of a node which forms part of the computer system in which an n by n array of nodes is a cluster, and bridging between the + and - x ports and the + and - y ports provide the cluster node interconnections, and each of the nodes has a plurality of said processor memory elements, each of which manages a single external port; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 140
Dependent Claim
252 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and wherein, processor memory element forms part of a node which forms part of the computer system in which an n by n array of nodes is a cluster, and bridging between the + and - x ports and the + and - y ports provide the cluster node interconnections, and each of the nodes has a plurality of said processor memory elements, each of which manages a single external port; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 141
Dependent Claim
219 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and wherein a processor memory element forms part of a node which has multiple processor memory elements and is connected to other nodes in a cluster network with I/O programs at dedicated interrupt levels for managing the network; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 142
Dependent Claim
219 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and wherein a processor memory element forms part of a node which has multiple processor memory elements and is connected to other nodes in a cluster network with I/O programs at dedicated interrupt levels for managing the network; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 143
Dependent Claim
276 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, the computer system organized as a massively parallel machine having nodes interconnected as a n dimensional network clusters with communication paths between said processor memory elements along said internal and external communication paths providing a processing array, each of the nodes including at least one of said processor memory elements, each of said clusters including an array of said processor memory elements, and the external communication paths forming edges of the clusters, and wherein loading or unloading the processing array is done by connecting the edges of at least one cluster to a system bus; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 144
Dependent Claim
276 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, the computer system organized as a massively parallel machine having nodes interconnected as a n dimensional network clusters with communication paths between said processor memory elements along said internal and external communication paths providing a processing array, each of the nodes including at least one of said processor memory elements, each of said clusters including an array of said processor memory elements, and the external communication paths forming edges of the clusters, and wherein loading or unloading the processing array is done by connecting the edges of at least one cluster to a system bus; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 145
Dependent Claim
264 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate wherein each of a plurality of nodes includes at least one of said processor memory elements and data transfers between two of said nodes occurs via a connection directly attached to a zipper interface, said zipper interface including means for selectively providing data transfer into and out of the two nodes, said zipper interface having a number of the connections to pairs of the nodes to satisfy application requirements for a given configuration of said processor memory elements, thereby providing different zipper sizes; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 146
Dependent Claim
264 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate wherein each of a plurality of nodes includes at least one of said processor memory elements and data transfers between two of said nodes occurs via a connection directly attached to a zipper interface, said zipper interface including means for selectively providing data transfer into and out of the two nodes, said zipper interface having a number of the connections to pairs of the nodes to satisfy application requirements for a given configuration of said processor memory elements, thereby providing different zipper sizes; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 147
Dependent Claim
277 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate wherein each of a plurality of nodes includes at least one of said processor memory elements, the nodes forming a node array which allows data transfers into or out of the node array to alternate between two of the nodes directly attached to a zipper interface into the node array, the computer system executing a program, said zipper interface having a width size that is varied depending on a given configuration of said processor memory elements, and depending on the program requirements to roll data and program into and out of the individual processor memory elements; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 148
Dependent Claim
277 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate wherein each of a plurality of nodes includes at least one of said processor memory elements, the nodes forming a node array which allows data transfers into or out of the node array to alternate between two of the nodes directly attached to a zipper interface into the node array, the computer system executing a program, said zipper interface having a width size that is varied depending on a given configuration of said processor memory elements, and depending on the program requirements to roll data and program into and out of the individual processor memory elements; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 149
Dependent Claim
240 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, said computer system including a zipper connection for a processor array of said processor memory elements, said zipper connection provided as a separate interface into said processor array, said zipper connection being established on one or two edges of a hypercube network by dotting onto a broadcast bus at multiple nodes in the processor array and in multiple directions; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 150
Dependent Claim
240 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, said computer system including a zipper connection for a processor array of said processor memory elements, said zipper connection provided as a separate interface into said processor array, said zipper connection being established on one or two edges of a hypercube network by dotting onto a broadcast bus at multiple nodes in the processor array and in multiple directions; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 151
Dependent Claim
190 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and wherein said common substrate is a CMOS chip; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 152
Dependent Claim
190 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, and wherein said common substrate is a CMOS chip; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 153
Dependent Claim
203 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, wherein the processor memory element further comprises multiple register sets for each interrupt level providing means for saving status when taking interrupts; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 154
Dependent Claim
203 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, wherein the processor memory element further comprises multiple register sets for each interrupt level providing means for saving status when taking interrupts; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 155
Dependent Claim
294 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, wherein each of a plurality of nodes includes a plurality of the processor memory elements, and the nodes are arranged in rings, each of the processor memory elements existing in only one ring of nodes, and wherein any of the processor memory elements selectively addresses as a neighbor any of the processor memory elements having and address that differs in a single bit position and is located in the same ring or an adjacent processor memory element within the same node, processor memory elements within the same node connected by additional rings such that communications is selectively routed between rings within the same node, thereby providing for myriad network topologies without pinout considerations; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 156
Dependent Claim
294 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, wherein each of a plurality of nodes includes a plurality of the processor memory elements, and the nodes are arranged in rings, each of the processor memory elements existing in only one ring of nodes, and wherein any of the processor memory elements selectively addresses as a neighbor any of the processor memory elements having and address that differs in a single bit position and is located in the same ring or an adjacent processor memory element within the same node, processor memory elements within the same node connected by additional rings such that communications is selectively routed between rings within the same node, thereby providing for myriad network topologies without pinout considerations; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 157
Dependent Claim
206 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, each processor memory element having an independent router function operative in the communication over said internal communication path and said dedicated, single external communication path; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
Claim 158
Dependent Claim
206 words
A computer system, comprising: a plurality of common substrates, each common substrate including a plurality of processor memory elements, each of said processor memory elements having a processor coupled to a dedicated local memory, each of said processor memory elements having an internal communication path for communication with any other of said plurality of processor memory elements located on the same common substrate, and each of said processor memory elements having a dedicated, single external communication path coupled directly to a respective one of said plurality of processor memory elements external of the same common substrate and located on another of said plurality of common substrates, for communication with other processor memory elements external to said substrate, each processor memory element having an independent router function operative in the communication over said internal communication path and said dedicated, single external communication path; and wherein said computer system has means on said substrate for distributing interconnection and controls within the system enabling the system to perform SIMD/MIMD functions as a multiprocessor memory system, wherein a local memory is dedicated to a processor and each dedicated local memory is independently accessible by the respectively coupled processor in both SIMD and MIMD modes exclusive of access by another processor.
References Cited (Backward Citations)
US3537074A
Cited
US3544973A
Cited
US4942516A
Cited
US3970993A
Cited
US4107773A
Cited
US4101960A
Cited
US4270170A
Cited
US4380046A
Cited
US4412303A
Cited
US4314349A
Cited
US4338675A
Cited
US4435758A
Cited
US4467422A
Cited
US4956772A
Cited
US5006978A
Cited
US4394726A
Cited
US4468727A
Cited
US4622650A
Cited
US4498133A
Cited
US4523273A
Cited
US4825359A
Cited
US4809159A
Cited
US4739474A
Cited
US4598400A
Cited
US5146608A
Cited
EP0132926B1
Cited
US5212773A
Cited
US5008815A
Cited
US4621339A
Cited
US4604695A
Cited
US4905143A
Cited
US4855903A
Cited
US4901224A
Cited
US5113523A
Cited
US4805091A
Cited
US4967340A
Cited
US5045995A
Cited
EP0208497B1
Cited
US5047917A
Cited
US4706191A
Cited
US4739476A
Cited
US4763321A
Cited
US5016163A
Cited
US4720780A
Cited
US4922408A
Cited
US4847755A
Cited
US5021945A
Cited
US4907148A
Cited
US4736291A
Cited
US4852048A
Cited
US4835729A
Cited
US4831519A
Cited
US4783782A
Cited
US4916657A
Cited
US4748585A
Cited
US4925311A
Cited
US5297260A
Cited
US4783738A
Cited
US4814980A
Cited
US4809169A
Cited
US4780873A
Cited
US4809347A
Cited
US4849882A
Cited
US5038386A
Cited
US4860201A
Cited
US4910665A
Cited
US4985832A
Cited
US5230079A
Cited
US5170484A
Cited
US4975834A
Cited
US4841476A
Cited
US5010477A
Cited
US4992933A
Cited
US5175865A
Cited
US5165023A
Cited
US4873626A
Cited
US4891787A
Cited
US4903260A
Cited
US4858110A
Cited
US4964032A
Cited
US4933846A
Cited
US4933895A
Cited
US5170482A
Cited
US5008882A
Cited
US4958273A
Cited
US4916652A
Cited
US4942517A
Cited
US4943912A
Cited
US4872133A
Cited
US4896265A
Cited
US4992926A
Cited
US5121498A
Cited
US5038282A
Cited
US5056000A
Cited
US5005120A
Cited
GB2223867A
Cited
US4920484A
Cited
US5179714A
Cited
US5041971A
Cited
US5189665A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2002-10-04 | 2003-04-05 | 2003-10-05 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2006-10-04 | 2007-04-05 | 2007-10-05 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2010-10-04 | 2011-04-05 | 2011-10-05 | $7700.00 | Window Closed / Lapsed |