Multiprocessor cache coherence management

Electrical computers and digital processing systems: memory – Address formation – Address mapping

Reexamination Certificate

Rate now

  [ 0.00 ] – not rated yet Voters 0   Comments 0

Details

C711S124000, C711S137000, C711S145000

Reexamination Certificate

active

06711662

ABSTRACT:

TECHNICAL FIELD
This invention relates to multiprocessor cache coherence management.
BACKGROUND
Referring to
FIG. 1
, a shared-memory multiprocessor system
10
includes processing modules
12
connected to an interconnection network
14
. Each processing module
12
includes a processor
121
and a cache
122
, which is a fast memory directly accessible to the associated processor
121
in the same processing module
12
. Cache
122
holds copies of data that have been recently accessed, and are likely to be accessed soon by its associated processor
121
. Before a processor
121
reads a data block, the processor first goes to its cache
122
to see if the data block has already been placed there. If the data block is not in its cache
122
, called a cache miss, or the data block is not valid, the processor must retrieve the data block from either a local or remote memory unit
16
through the interconnection network
14
. The interconnection network
14
is typically a bus or a general Local Area Network (LAN) that delivers data to its destination according to a destination address sent with the data. An I/O controller
18
, also connected to the interconnection network
14
, serves as an I/O interface to various types of I/O devices.
The multiprocessor system
10
includes memory units
16
, each coupled to, or associated with, one of the processing modules
12
. The memory units
16
are shared by all of the processors
121
, that is, every processor
121
can read from or write to any of memory units
16
. However, only the processor
121
associated with, i.e., locally connected to, a memory unit
16
has local access to that memory unit; all the other processors
121
have to access it remotely through the interconnection network
14
.


REFERENCES:
patent: 5752258 (1998-05-01), Guzovskiy et al.
patent: 6052760 (2000-04-01), Bauman et al.
patent: 6108737 (2000-08-01), Sharma et al.
patent: 6151663 (2000-11-01), Pawlowski et al.
patent: 6341334 (2002-01-01), Kamemaru
patent: 6449696 (2002-09-01), Okayasu
Markatos, Evangelos P., et al., “A Top-10 Approach to Prefetching on the Web”, Institute for Computer Science, Aug. 1996.*
Alternative Implementations of Two-Level Adaptive Branch Predictionby Tse-yu Yeh and Yale N. Patt, Department of Electrical Engineering and Computer Science, University of Michigan; published at the 19thAnnual International Symposium on Computer Science.

LandOfFree

Say what you really think

Search LandOfFree.com for the USA inventors and patents. Rate them and share your experience with other people.

Rating

Multiprocessor cache coherence management does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Multiprocessor cache coherence management, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Multiprocessor cache coherence management will most certainly appreciate the feedback.

Rate now

     

Profile ID: LFUS-PAI-O-3251901

  Search
All data on this website is collected from public sources. Our data reflects the most accurate information available at the time of publication.