Patent US533597A successfully refreshed from Google Patents.
Patent US1923065A successfully refreshed from Google Patents.
Patent US5754877A successfully refreshed from Google Patents.
← Back to Patent Explorer
US5754877A Expired / Lapsed US ⏱️ 0.0 years remaining

Extended symmetrical multiprocessor architecture

Assignee: Oracle America Inc
•
Filing: 1996-07-02
•
Granted: 1998-05-19
•
Expires: 2016-07-02
📄 Original PDF 🌐 Google Patents 🔄 Sync Now

Abstract

Abstract An architecture for an extended multiprocessor (XMP) computer system is provided. The XMP computer system includes multiple SMP nodes. Each SMP node includes an XMP interface and a repeater structure coupled to the XMP interface. The SMP nodes are connected to each other by unidirectional point-to-point links. The repeater structure in each SMP node includes an upper level bus, one or more transaction repeaters coupled to the upper level bus. Each transaction repeater broadcasts transactions to bus devices attached to a lower level bus, wherein each transaction repeater is coupled to a separate lower level bus. Transaction repeater includes a queue and a bypass path. Transaction originating in a particular SMP node are stored in the queue, whereas transactions originating in other SMP nodes bypass the incoming queue to the bus device. Multiple transactions may be simultaneously broadcast across the point-to-point link connections between the SMP nodes. However, transactions are broadcast to the SMP nodes in a defined, uniform order. A control signal is asserted by the XMP interface so that a transaction is received by bus devices in the originating node from the incoming queues at the same time and in the same order it is received by bus devices in non-originating nodes. Thus a hierarchical bus structure is provided that overcomes physical/electrical limitations of single bus architecture while maximizing bus bandwidth utilization.

Inventors

  • 👤 Erik E. Hagersten
  • 👤 Mark D. Hill

Bibliographic Details

Application Number —
Jurisdiction US
Data Source google_patents
Last Synchronized 2026-10-28 12:14
CPC Classifications
G; G06; G06F; G06F15/00; G06F15/16; G06F15/163; G06F15/173; G06F15/17356; G06F15/17368; G06F15/17381

Claims Analysis

Total Claims: 18
Claim 1
Independent Claim 129 words
A node in a multiprocessor computer system, comprising: a top level interface that receives incoming transactions and transmits outgoing transactions, wherein said outgoing transactions originate in said node and said incoming transaction do not originate in said node; an upper level bus; a first repeater, wherein said first repeater is coupled to said top level interface by said upper level bus, and wherein said first repeater includes an incoming queue and a bypass path; wherein said first repeater receives said incoming transactions from said top level interface and transmits said incoming transactions via said bypass path to a lower level bus, and wherein said first repeater receives said outgoing transaction on said upper level bus and transmits said outgoing transactions via said incoming queue to a lower level bus.
Claim 2
Dependent Claim 129 words
A node in a multiprocessor computer system, comprising: a top level interface that receives incoming transactions and transmits outgoing transactions, wherein said outgoing transactions originate in said node and said incoming transaction do not originate in said node; an upper level bus; a first repeater, wherein said first repeater is coupled to said top level interface by said upper level bus, and wherein said first repeater includes an incoming queue and a bypass path; wherein said first repeater receives said incoming transactions from said top level interface and transmits said incoming transactions via said bypass path to a lower level bus, and wherein said first repeater receives said outgoing transaction on said upper level bus and transmits said outgoing transactions via said incoming queue to a lower level bus.
Claim 3
Dependent Claim 43 words
The node as recited in claim 1, wherein said top level interface receives said incoming transactions on an unidirectional point-to-point link with another node, and wherein said top level interface transmits said outgoing transactions on an unidirectional point-to-point link with said another node.
Claim 4
Dependent Claim 43 words
The node as recited in claim 1, wherein said top level interface generates a control signal received by said first repeater for controlling when said first repeater transmits transactions to said lower level bus via said bypass path or via said incoming queue.
Claim 5
Dependent Claim 51 words
The node as recited in claim 1, wherein said first repeater transmits said outgoing transactions via said incoming queue to said lower level bus at approximately the same time as each of said outgoing transactions are transmitted on a corresponding lower level bus of another node in the multiprocessor computer system.
Claim 6
Dependent Claim 50 words
The node as recited in claim 1, wherein said first repeater transmits said outgoing transactions via said incoming queue to said lower level bus in the same order as each of said outgoing transactions are transmitted on a corresponding lower level bus of another node in the multiprocessor computer system.
Claim 7
Dependent Claim 48 words
The node as recited in claim 1, further comprising a processor device coupled to said lower level bus, wherein said processor device is configured to receive transactions transmitted on said lower level bus by said repeater, and wherein said processor device is configured to initiate said outgoing transactions.
Claim 8
Dependent Claim 27 words
The node as recited in claim 6, wherein said processor device further comprises a SPARC processor and a level two cache memory coupled to said SPARC processor.
Claim 9
Dependent Claim 37 words
The node as recited in claim 6, further comprising an input/output (I/O) device coupled to said lower level bus, wherein said I/O device is configured to receive transactions transmitted on said lower level bus by said repeater.
Claim 10
Dependent Claim 51 words
The node as recited in claim 1, further comprising a second repeater, wherein said second repeater is coupled to said top level interface by said upper level bus and coupled to a second lower level bus, and wherein said second repeater includes a second incoming queue and a second bypass path.
Claim 11
Dependent Claim 35 words
The node as recited in claim 9, wherein said second repeater receives said incoming transactions from said top level interface and transmits said incoming transactions via said bypass path to said second lower level bus.
Claim 12
Dependent Claim 159 words
A multiprocessor computer system, comprising: a first multiprocessor node; a second multiprocessor node coupled to said first multiprocessor node by two unidirectional transaction links, wherein said two unidirectional transaction links are configured to transmit transactions from said first multiprocessor node to said second multiprocessor node while simultaneously transmitting transactions from said second multiprocessor node to said first multiprocessor node; wherein said first multiprocessor node comprises: a first top level interface coupled to said two unidirectional transaction links; a first upper level bus; a first repeater coupled to said first top level interface by said first upper level bus, wherein said first repeater broadcasts transactions from said first upper level bus to a first lower level bus via a bypass path in said first repeater when said transactions originated from said second multiprocessor node, and wherein said first repeater broadcasts transactions to said first lower level bus via an incoming queue when said transactions originated in said first multiprocessor node.
Claim 13
Dependent Claim 159 words
A multiprocessor computer system, comprising: a first multiprocessor node; a second multiprocessor node coupled to said first multiprocessor node by two unidirectional transaction links, wherein said two unidirectional transaction links are configured to transmit transactions from said first multiprocessor node to said second multiprocessor node while simultaneously transmitting transactions from said second multiprocessor node to said first multiprocessor node; wherein said first multiprocessor node comprises: a first top level interface coupled to said two unidirectional transaction links; a first upper level bus; a first repeater coupled to said first top level interface by said first upper level bus, wherein said first repeater broadcasts transactions from said first upper level bus to a first lower level bus via a bypass path in said first repeater when said transactions originated from said second multiprocessor node, and wherein said first repeater broadcasts transactions to said first lower level bus via an incoming queue when said transactions originated in said first multiprocessor node.
Claim 14
Dependent Claim 46 words
The multiprocessor computer system of claim 11, wherein said first top level interface generates a first control signal received by said first repeater for controlling when said first repeater transmits transactions to said first lower level bus via said bypass path or via said incoming queue.
Claim 15
Dependent Claim 27 words
The multiprocessor computer system of claim 12, wherein said transactions that originate in said first multiprocessor node are stored in said incoming queue in said first repeater.
Claim 16
Dependent Claim 56 words
The multiprocessor computer system of claim 13, wherein said second multiprocessor node comprises: a second top level interface coupled to said two unidirectional transaction links; a second upper level bus; a second repeater coupled to said second top level interface by said second upper level bus; a second lower level bus coupled to said second repeater.
Claim 17
Dependent Claim 154 words
The multiprocessor computer system of claim 14, wherein said transactions originating in said first multiprocessor node are received by said second top level interface and transmitted over said second upper level bus to said second repeater, wherein said second repeater broadcasts said transactions originating in said first multiprocessor node to said second lower level bus; wherein said first top level interface asserts said first control signal so that said first repeater broadcasts said transactions originating in said first multiprocessor node to said first lower level bus via said incoming queue at approximately the same time as each of said transactions originating from first multiprocessor node are broadcast on said second lower level bus; and wherein new transactions originating in said first multiprocessor node are transmitted on said first upper level bus at the same time previous ones of said transactions originating in said first multiprocessor node are transmitted over said second upper level bus.
Claim 18
Dependent Claim 87 words
The multiprocessor computer system of claim 11, wherein said second multiprocessor node comprises: a second top level interface coupled to said two unidirectional transaction links; a second upper level bus; a second repeater coupled to said second top level interface by said second upper level bus; a second lower level bus coupled to said second repeater; wherein all transactions appear at most only once on each of said first lower level bus, said second lower level bus, said first upper bus, and said second upper level bus.

References Cited (Backward Citations)

US4819232A Cited
US5608893A Cited
US5590335A Cited
US5533103A Cited
US5579512A Cited
US5655103A Cited

USPTO Annuity & Maintenance Windows

Annuity Stage Window Opens Payment Deadline Surcharge Period Until Estimated Fee Status
3.5 Years (Year 4) 2001-05-18 2001-11-17 2002-05-19 $2000.00 Window Closed / Lapsed
7.5 Years (Year 8) 2005-05-18 2005-11-17 2006-05-19 $3760.00 Window Closed / Lapsed
11.5 Years (Year 12) 2009-05-18 2009-11-17 2010-05-19 $7700.00 Window Closed / Lapsed

Assigned Categories & Tags

No categories assigned to this patent yet.

Assign Category Tag