US4814970A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Multiple-hierarchical-level multiprocessor system
Assignee: Elsag Bailey SpA
•
Filing: 1986-12-08
•
Granted: 1989-03-21
•
Expires: 2006-12-08
Abstract
Abstract A multiple-hierarchical-level multiprocessor system having first numbers of processing modules comprising, each, at least one processor and connected to a first group of common direct-access communication lines for forming a first hierarchical level (family); a first processing module of each of the aforementioned first numbers also being connected to a second group of common direct-access communication lines for forming a second hierarchical level (region); a second number of the aforementioned second groups of common communication lines being interconnected via data transmitting and receiving means for forming a third hierarcical level (region network); the aforementioned means comprising at least one processor for enabling operation independent of the aforementioned modules.
Inventors
- 👤 Giuseppe Barbagelata
- 👤 Bruno Conterno
- 👤 Vildo Luperini
- 👤 Enrico Perroni
- 👤 Fernando Pesce
- 👤 Osvaldo Pugliese
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-24 18:54
CPC Classifications
G; G06; G06F; G06F13/00; G06F13/38; G06F13/40; G06F13/4004; G06F13/4027; G06F15/00; G06F15/16; G06F15/163; G06F15/76; G06F15/80; G06F15/8007
Claims Analysis
Total Claims: 29
Claim 1
Independent Claim
163 words
A multiple-hierarchical-level multiprocessor system, comprising a plurality of regions (10) of families (11) of processor modules (17, 18), each of said modules comprising at least one processor (31, 31'), and each of said modules in each of said families being connected to a first common direct access communication line (12) for each of said families, the family of modules so connected thereby constituting a first hierarchical-level; wherein a first processing module (17) of each of said families (11) in each of said regions (10) is also connected to a second common direct-access communication line (13), the region of families so connected thereby forming a second hierarchical level; and wherein a plurality of said second lines (13) are connected to one another by data transmitting and receiving means (14), the region network so connected thereby forming a third hierarchical level; said plurality of transmitting and receiving means (14) each comprising at least one processor (31") enabling independent operation of said processor modules (17, 18).
Claim 2
Dependent Claim
163 words
A multiple-hierarchical-level multiprocessor system, comprising a plurality of regions (10) of families (11) of processor modules (17, 18), each of said modules comprising at least one processor (31, 31'), and each of said modules in each of said families being connected to a first common direct access communication line (12) for each of said families, the family of modules so connected thereby constituting a first hierarchical-level; wherein a first processing module (17) of each of said families (11) in each of said regions (10) is also connected to a second common direct-access communication line (13), the region of families so connected thereby forming a second hierarchical level; and wherein a plurality of said second lines (13) are connected to one another by data transmitting and receiving means (14), the region network so connected thereby forming a third hierarchical level; said plurality of transmitting and receiving means (14) each comprising at least one processor (31") enabling independent operation of said processor modules (17, 18).
Claim 3
Dependent Claim
96 words
The invention according to claim 1, wherein each of said processing modules (17, 18) comprises, assigned to each processor (31, 31'): a common private-level communication line (32, 32'), and private memory means (34, 34') and private input-output means (37, 37') connected to said common private-level communication line (32, 32'); a plurality of common local-level communication lines (51, 51'); and local memory means (63, 63') connected to said local-level communication lines (51, 51') and to said private input-output means (37, 37'); wherein said private input-output means (37, 37') is separable from said local-level communication line (51, 51').
Claim 4
Dependent Claim
21 words
The invention according to claim 2, further comprising coprocessors (33, 33') assigned to at least one of said processors (31, 31').
Claim 5
Dependent Claim
43 words
The invention according to claim 2, wherein said private input-output means (37, 37') comprises timing means (40, 40'), control (38, 38') and status (39, 39') registers, and interrupt signal control means (41, 41'), each connected to said common private-level communication lines (32, 32').
Claim 6
Dependent Claim
38 words
The invention according to claim 2, wherein said processing modules (18) other than said first processing modules (17) are connected solely to said first direct-access communication line (12), and wherein each comprises a plurality of said processors (31').
Claim 7
Dependent Claim
67 words
The invention according to claim 2, wherein said processing modules (17, 18) each comprise means for detecting requests for connecting said local-level communication lines (51, 51') to said private-level communication lines (32, 32') or to said first (12) or second (13) direct-access communication lines, and for enabling direct-access connection of said communication lines (12, 13, 32, 32', 51, 51') in accordance with an associated request priority sequence.
Claim 8
Dependent Claim
49 words
The invention according to claim 1, further comprising means (67, 68) for sending data from one of said processor modules (17, 18) to at least one other processor module (17, 18) connected to the same group of direct-access communication lines (12, 13) as said one processor module (17, 18).
Claim 9
Dependent Claim
77 words
(New) The invention according to claim 1, wherein said data transmitting and receiving means (14) further comprises first means (98) for temporarily storing data for transmission, second means (99) for temporarily storing received data, and means (101, 102, 103, 104, 105) addressing said first (98) and second (99) means for controlling exchange of said data; said control means (101, 102, 103, 104, 105) being controlled by said processor (31") of said data transmitting and receiving means (14).
Claim 10
Dependent Claim
110 words
The invention according to claim 8, wherein said data transmitting and receiving means (14) comprises: a common private-level communication line (32"), and private memory means (34") and private input-output means (37") connected to said common private-level communication line (32"); a plurality of common local-level communication lines (51"); and local memory means (63") and local input-output means connected to each of said local-level communication lines (51"); wherein said private memory (34") and input-output (37") means are separable from said local-level communication lines (51"); and wherein said system further comprises interface means (57") for connecting said local-level communication lines (51") to said second direct-access communication lines (13) at said second hierarchical level.
Claim 11
Dependent Claim
25 words
The invention according to claim 1, wherein said data transmitting and receiving means (14) comprises TTL-differential and differential-TTL signal converting means (95, 96, 95', 96').
Claim 12
Dependent Claim
90 words
The invention according to claim 1, further comprising a second main regional group (13') of common communication lines at said second hierarchical level, wherein at least one of said common communication direct-access lines (13) are connected to said lines of said second main regional group (13') by a subset of said data transmitting and receiving means (14); and further comprising external processing means (16) to which said lines of said second main regional group (13') are connected by others of said transmitting and receiving means (14) not of said subset.
Claim 13
Dependent Claim
33 words
The invention according to claim 1, further comprising at least one memory means (22) connected in direct access manner to both said first (12) and said second (13) groups of common communication lines.
Claim 14
Dependent Claim
41 words
The invention according to claim 12, wherein said memory means (22) offers a high addressing capacity in access of the addressing capacity of said processing modules (17, 18) connected to said first (12) and said second (13) group of communication lines.
Claim 15
Dependent Claim
102 words
The invention according to claim 12, wherein, for each group of address data from said first (12) and said second (13) group of common communication lines, said memory means (22) comprises match coding means (168, 168') for a first part (165, 165') of each of said groups of address data, the remainder (173, 173') of each of said groups of address data being reunited with said part obtained from said match coding means (168, 168'); and wherein said system further comprises selecting means (161, 182) for selecting a complete group of said address data to be supplied to a memory block (150).
Claim 16
Dependent Claim
30 words
The invention according to claim 14, wherein said first part of said group of said address data sent to said match coding means (168, 168') is the most significant part.
Claim 17
Dependent Claim
17 words
The invention according to claim 14, wherein said match coding means (168, 168') comprises read-and-write memory means.
Claim 18
Dependent Claim
37 words
The invention according to claim 14, wherein read or write data from said memory block (150) is exchanged on said first (12) or said second (13) group of common communication lines by registers (154, 156; 155, 157).
Claim 19
Dependent Claim
77 words
The invention according to claim 1, wherein said first (12) and said second (13) group of communication lines and said data transmitting and receiving means (14) are adapted to provide a data exchange parallelism which is a multiple of the word size of said processors (31, 31', 31"); and wherein said system further comprises interface means (63, 261, 250) between said processors (31, 31', 31") and said common communication lines (12, 13) for achieving said multiple parallelism.
Claim 20
Dependent Claim
22 words
The invention according to claim 18, wherein said data exchange parallelism is twice the word size of said processors (31, 32', 31").
Claim 21
Dependent Claim
107 words
The invention according to claim 18, wherein said interface means comprises a data exchange part (261) in said common local-level communication line (51) providing said multiple parallelism; wherein said data exchange part (261) comprises lines connected to a plurality of memory banks (257, 258, 259, 260) connected to address lines (249) from said processors (31, 31', 31") through address altering means (250) or from external processing modules (17, 18, 14); and wherein said address altering means (250) is deactivated when the parallelism of said data exchanged between said processing modules is the same as or a submultiple of said word size of said processors (31, 31', 31").
Claim 22
Dependent Claim
58 words
The invention according to claim 20, further comprising a plurality of direct lines connected to single elements of said memory banks (257, 258, 259, 260), and a plurality of address lines connected to said logic blocks (256); wherein said single elements are addressed in response to both said direct lines and signals derived from said logic blocks (256).
Claim 23
Dependent Claim
66 words
The invention according to claim 20, wherein any of said data exchange lines (261) of said common local-level communication lines (51) which exceed the word size of said processors (31, 31', 31") are interconnected by temporary memory means (271), and are connected by interconnecting means (272) to the part of said data exchange lines (261) included within the word length of said processors (31, 31', 31").
Claim 24
Dependent Claim
96 words
The invention according to claim 1, further comprising means (54, 31, 40, 38, 57) for enabling single communication exchange between all of said modules (17, 18, 14) connected to the same first (12) and second (13) group of common communication lines and which emit an exchange request signal (FPREN) within a predetermined time period; and for enabling a locked multiple communication exchange by only an authorized one of said modules (17, 18, 14), so as to substantially enable forecasting of both an interrupt signal wait time and an exchange time for a given set of data.
Claim 25
Dependent Claim
58 words
The invention according to claim 23, further comprising interfaces (54, 54', 57, 57") connecting said modules (17, 18, 14) to said first (12) or said second (13) group of common communication lines, wherein said single and locked multiple communication exchange enabling means (54, 31, 40, 38, 57) form at least part of said interfaces (54, 54', 57, 57").
Claim 26
Dependent Claim
47 words
The invention according to claim 24, wherein said single and locked multiple communication enabling means (54, 31, 40, 38, 57) are controlled by said processors (31, 31', 31") by the status of internal registers (38, 38', 38"), and by timing means (40, 40', 40"), in said modules.
Claim 27
Dependent Claim
51 words
(New) The invention according to claim 23, wherein said single and locked multiple common communication exchange enabling means (54, 31, 40, 38, 57) of each of said modules (17, 18, 14) are connected in daisy chain fashion, and wherein said first one (17) of said modules comprises initializing means (299, 85).
Claim 28
Dependent Claim
65 words
(New) The invention according to claim 2, further comprising means (47, 47', 47") for each of said processing modules (17, 18, 14) for decoding requests from said processor (31, 31', 31") for direct access to private, local, family or region areas of said memory means (34, 63, 63', 63"), and of said input-output means of said processing modules (17, 18) or said interconnecting means (14).
Claim 29
Dependent Claim
94 words
The invention according to claim 1, wherein a few of said processors (31') of each of said processing modules (18) are connected solely to said first (12) group of said common communication lines, wherein a few tens of said processing modules (18) are connected solely to said first (12) group of lines, and wherein a few tens of families (11) of processing modules (17, 18) are connected to said second (13) group of common communication lines, and a few regions (10) of families (11) are interconnected by said data transmitting and receiving means (14).
References Cited (Backward Citations)
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1992-03-20 | 1992-09-19 | 1993-03-21 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 1996-03-20 | 1996-09-19 | 1997-03-21 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2000-03-20 | 2000-09-19 | 2001-03-21 | $7700.00 | Window Closed / Lapsed |