Patent US6108764A successfully refreshed from Google Patents.
US6108764A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Non-uniform memory access (NUMA) data processing system with multiple caches concurrently holding data in a recent state from which data can be sourced by shared intervention
Assignee: International Business Machines Corp
•
Filing: 1998-12-17
•
Granted: 2000-08-22
•
Expires: 2018-12-17
Abstract
Abstract A non-uniform memory access (NUMA) computer system includes first and second processing nodes that are coupled together. The first processing node includes a system memory and first and second processors that each have a respective associated cache hierarchy. The second processing node includes at least a third processor and a system memory. If the cache hierarchy of the first processor holds an unmodified copy of a cache line and receives a request for the cache line from the third processor, the cache hierarchy of the first processor sources the requested cache line to the third processor and retains a copy of the cache line in a Recent coherency state from which the cache hierarchy of the first processor can source the cache line in response to subsequent requests.
Inventors
- 👤 Yoanna Baumgartner
- 👤 Anna Elman
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-08 15:28
CPC Classifications
G; G06; G06F; G06F12/00; G06F12/02; G06F12/08; G06F12/0802; G06F12/0806; G06F12/0813
Claims Analysis
Total Claims: 12
Claim 1
Independent Claim
163 words
A computer system, comprising: first and second processing nodes coupled together, wherein said first processing node includes a system memory and a first processor having a first cache hierarchy and a second processor having a second cache hierarchy, and wherein said second processing node includes at least a system memory and a third processor having a third cache hierarchy; wherein said first cache hierarchy, responsive to receipt of a read request from the second processing node for a copy of an unmodified cache line held by said first cache hierarchy, sources a copy of said unmodified cache line to said third cache hierarchy and retains said unmodified cache line in a Recent coherency state from which said first cache hierarchy can source said cache line; and wherein said third cache hierarchy, responsive to receipt of said copy of said unmodified cache line, concurrently stores said cache line in said Recent state to permit said third cache hierarchy to also source said cache line.
Claim 2
Dependent Claim
163 words
A computer system, comprising: first and second processing nodes coupled together, wherein said first processing node includes a system memory and a first processor having a first cache hierarchy and a second processor having a second cache hierarchy, and wherein said second processing node includes at least a system memory and a third processor having a third cache hierarchy; wherein said first cache hierarchy, responsive to receipt of a read request from the second processing node for a copy of an unmodified cache line held by said first cache hierarchy, sources a copy of said unmodified cache line to said third cache hierarchy and retains said unmodified cache line in a Recent coherency state from which said first cache hierarchy can source said cache line; and wherein said third cache hierarchy, responsive to receipt of said copy of said unmodified cache line, concurrently stores said cache line in said Recent state to permit said third cache hierarchy to also source said cache line.
Claim 3
Dependent Claim
65 words
The computer system of claim 1, wherein said first cache hierarchy, responsive to receipt of a request by said third cache hierarchy for a copy of a cache line that said first cache hierarchy holds in modified state, said first cache hierarchy sources a copy of said modified cache line to said third cache hierarchy and retains said cache line in said Recent coherency state.
Claim 4
Dependent Claim
37 words
The computer system of claim 1, wherein said first cache hierarchy retains said cache line in said Recent coherency state only in response to receipt of an indication that said read request is from another processing node.
Claim 5
Dependent Claim
34 words
The computer system of claim 1, wherein prior to receipt of said read request, said first cache hierarchy associates said unmodified cache line with a coherency state that is one of Exclusive and Recent.
Claim 6
Dependent Claim
57 words
The computer system of claim 3, wherein said first and second processing nodes are coupled by a node interconnect and each contain a respective one of first and second node controllers for managing communication over said interconnect, wherein one of said first and second node controllers provides said indication in response to receipt of said read request.
Claim 7
Independent Claim
176 words
A method of data sharing in a computer system including first and second processing nodes coupled together, wherein said first processing node includes a system memory and a first processor having a first cache hierarchy and a second processor having a second cache hierarchy, and wherein said second processing node includes at least a system memory and a third processor having a third cache hierarchy, said method comprising: at said first cache hierarchy, receiving from said third processor a read request for a cache line; and if said requested cache line is held in said first cache hierarchy in an unmodified state, sourcing a copy of said unmodified cache line from said first cache hierarchy to said third cache hierarchy, retaining said cache line in said first cache hierarchy in a Recent coherency state from which said first cache hierarchy can source said cache line, and concurrently storing said copy of said cache line in said third cache hierarchy in said Recent coherency state to also permit said third cache hierarchy to source said cache line.
Claim 8
Independent Claim
176 words
A method of data sharing in a computer system including first and second processing nodes coupled together, wherein said first processing node includes a system memory and a first processor having a first cache hierarchy and a second processor having a second cache hierarchy, and wherein said second processing node includes at least a system memory and a third processor having a third cache hierarchy, said method comprising: at said first cache hierarchy, receiving from said third processor a read request for a cache line; and if said requested cache line is held in said first cache hierarchy in an unmodified state, sourcing a copy of said unmodified cache line from said first cache hierarchy to said third cache hierarchy, retaining said cache line in said first cache hierarchy in a Recent coherency state from which said first cache hierarchy can source said cache line, and concurrently storing said copy of said cache line in said third cache hierarchy in said Recent coherency state to also permit said third cache hierarchy to source said cache line.
Claim 9
Dependent Claim
56 words
The method of claim 6, and further comprising: if said requested cache line is held in said first cache hierarchy in a modified state, sourcing a copy of said cache line from said first cache hierarchy to said third cache hierarchy and retaining said cache line in said first cache hierarchy in said Recent coherency state.
Claim 10
Dependent Claim
42 words
The method of claim 6, wherein retaining said cache line in said Recent coherency state comprises retaining said cache line in said Recent coherency state only in response to receipt of an indication that said read request is from another processing node.
Claim 11
Dependent Claim
36 words
The method of claim 6, and further comprising: prior to receipt of said read request, associating said unmodified cache line with a coherency state that is one of Exclusive and Recent in said first cache hierarchy.
Claim 12
Dependent Claim
60 words
The method of claim 9, wherein said first and second processing nodes are coupled by a node interconnect and each contain a respective one of first and second node controllers for managing communication over said interconnect, said method further comprising generating said indication at one of said first and second node controllers in response to receipt of said read request.
References Cited (Backward Citations)
US5673413A
Cited
US5893144A
Cited
US5878268A
Cited
US5940864A
Cited
US5940856A
Cited
US5943685A
Cited
US5943684A
Cited
US5963974A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2003-08-22 | 2004-02-21 | 2004-08-22 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2007-08-22 | 2008-02-21 | 2008-08-22 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2011-08-22 | 2012-02-21 | 2012-08-22 | $7700.00 | Window Closed / Lapsed |