Work Files Saved Searches
   My Account                                                  Search:   Quick/Number   Boolean   Advanced   Derwent    Help   


 The Delphion Integrated View

  Buy Now:   Buy PDF- 21pp  PDF  |   File History  |   Other choices   
  Tools:  Citation Link  |  Add to Work File:    
  View:  Expand Details   |  INPADOC   |  Jump to: 
  Go to:  Derwent  
 Email this to a friend  Email this to a friend 
       
Title: US7225300: Duplicate snoop tags partitioned across multiple processor/cache chips in a multi-processor system
[ Derwent Title ]


Country: US United States of America

View Images High
Resolution

 Low
 Resolution

 
21 pages

 
Inventor: Choquette, Jack H.; Mountain View, CA, United States of America
Kruckemyer, David A.; Mountain View, CA, United States of America
Hathaway, Robert G.; Sunnyvale, CA, United States of America

Assignee: Azul Systems, Inc, Mountain View, CA, United States of America
other patents from AZUL SYSTEMS, INC. (874053) (approx. 2)
 News, Profiles, Stocks and More about this company

Published / Filed: 2007-05-29 / 2004-09-15

Application Number: US2004000711387

IPC Code: Advanced: G06F 12/00; G06F 13/00;
Core: more...

ECLA Code: G06F12/08B4P4;

U.S. Class: 711/146; 711/140; 711/147; 711/119;

Field of Search: Non/00e

Priority Number:
2004-09-15  US2004000711387

Abstract:     Several cluster chips and a shared main memory are connected by interconnect buses. Each cluster chip has multiple processors using multiple level-2 local caches, two memory controllers and two snoop tag partitions. The interconnect buses connect all local caches to all snoop tag partitions on all cluster chips. Each snoop tag partition has all the system's snoop tags for a partition of the main memory space. The snoop index is a subset of the cache index, with remaining chip-select and interleave address bits selecting which of the snoop tag partitions on the multiple cluster chips stores snoop tags for that address. The number of snoop entries in a snoop set is equal to a total number of cache entries in one cache index for all local caches on all cluster chips. Cache coherency request processing is distributed among the snoop tag partitions on different cluster chips, reducing bottlenecks.

Attorney, Agent or Firm: g Patent LLC ; Auvinen, Stuart T. ;

Primary / Asst. Examiners: Sough, Hyung; Chery, Mardochee

INPADOC Legal Status: Show legal status actions

Family: None

First Claim:
Show all 20 claims
    1. A multi-processor system comprising:

a plurality of snoop tag partitions for storing snoop entries;

a plurality of external interconnect buses coupled between a plurality of snoop tag partitions, the plurality of external interconnect buses carrying cache coherency requests that include a snoop address;

a plurality of memory controllers coupled to a shared main memory;

a plurality of local processors for executing instructions and reading and writing data;

a plurality of local caches, coupled to the plurality of local processors, for storing cache entries that contain instructions or data used by the plurality of processors;

internal interconnect buses that couple the plurality of snoop tag partitions to the plurality of local caches and to the plurality of external interconnect buses;

wherein each snoop tag partition in the plurality of snoop tag partitions contains snoop entries arranged into snoop sets, wherein a snoop index selects one of the snoop sets as a selected snoop set, wherein all snoop entries within a snoop set have a same snoop index but are able to have different snoop tags;

wherein each local cache in the plurality of local caches contain cache entries arranged as multi-way cache sets, wherein a cache index selects one of the cache sets as a selected cache set, wherein all cache entries within a cache set have a same cache index but have different cache tags;

wherein the snoop address carried over the internal interconnect buses comprises a tag portion for matching with a cache tag, a cache-index portion having the cache index for selecting the selected cache set, and an offset portion of data within a selected cache entry, wherein the cache-index portion further comprises a snoop-index portion having the snoop index for selecting the selected snoop set, a chip-select portion, and an interleave portion;

wherein the chip-select portion of the cache-index portion of the snoop address selects a selected group of snoop tag partitions in the plurality of snoop tag partitions;

wherein the interleave portion of the cache-index portion of the snoop address selects a selected snoop tag partition in the plurality of snoop tag partitions within the selected group of snoop tag partitions;

wherein the selected snoop tag partition responds to the cache coherency request having the snoop address and stores a snoop tag in a snoop entry within the selected snoop set selected by the snoop index;

wherein other snoop tag partitions do not respond to the cache coherency request,

wherein the selected snoop tag partition is selected by the chip-select portion and the interleave portion of the snoop address which are subsets of the cache index,

whereby processing of snoop requests are partitioned across the plurality of snoop tag partitions by the chip-select portion of the snoop address.



Background / Summary: Show background / summary

Drawing Descriptions: Show drawing descriptions

Description: Show description

       
U.S. References: Go to Result Set: All U.S. references   |  No patents reference this one   |   Backward references (12)   |   Citation Link

Buy
PDF
Patent  Pub.Date  Inventor Assignee   Title
Buy PDF- 11pp US5590310  1996-12 Willenz et al.  Integrated Device Technology, Inc. Method and structure for data integrity in a multiple level cache system
Buy PDF- 70pp US5634068  1997-05 Nishtala et al.  Sun Microsystems, Inc. Packet switched cache coherent multiprocessor system
Buy PDF- 13pp US5765196  1998-06 Liencres et al.  Sun Microsystems, Inc. System and method for servicing copyback requests in a multiprocessor system with a shared memory
Buy PDF- 9pp US5946709  1999-08 Arimilli et al.  International Business Machines Corporation Shared intervention protocol for SMP bus using caches, snooping, tags and prioritizing
Buy PDF- 8pp US5978886  1999-11 Moncton et al.  Hewlett-Packard Company Method and apparatus for duplicating tag systems to maintain addresses of CPU data stored in write buffers external to a cache
Buy PDF- 9pp US6076147  2000-06 Lynch et al.  Sun Microsystems, Inc. Non-inclusive cache system using pipelined snoop bus
Buy PDF- 73pp US6108752  2000-08 VanDoren et al.  Compaq Computer Corporation Method and apparatus for delaying victim writes in a switch-based multi-processor system to maintain data coherency
Buy PDF- 16pp US6272602  2001-08 Singhal et al.  Sun Microsystems, Inc. Multiprocessing system employing pending tags to maintain cache coherence
Buy PDF- 13pp US6493797  2002-12 Lee et al.  Intel Corporation Multi-tag system and method for cache read/write
Buy PDF- 55pp US20020087807A1  2002-07 Gharachorloo et al.   System for minimizing directory information in scalable multiprocessor systems with logically independent input/output nodes
Buy PDF- 12pp US20020174305A1  2002-11 Vartti   METHOD AND APPARATUS FOR CONTROLLING MEMORY STORAGE LOCKS BASED ON CACHE LINE OWNERSHIP
Buy PDF- 22pp US20040068616A1  2004-04 Tierney et al.   System and method enabling efficient cache line reuse in a computer system
       
Foreign References: None

Inquire Regarding Licensing

Powered by Verity


Plaques from Patent Awards      Gallery of Obscure PatentsNominate this for the Gallery...

Thomson Reuters Copyright © 1997-2010 Thomson Reuters 
Subscriptions  |  Web Seminars  |  Privacy  |  Terms & Conditions  |  Site Map  |  Contact Us  |  Help