Abstract | ||
---|---|---|
The M-Machine is an experimental multicomputer being developed to test architectural concepts motivated by the constraints
of modern semiconductor technology and the demands of programming systems. The M-Machine computing nodes are connected with
a 3-D mesh network; each node is a multithreaded processor incorporating 9 function units, on-chip cache, and local memory.
The multiple function units are used to exploit both instruction-level and thread-level parallelism. A user accessible message
passing system yields fast communication and synchronization between nodes. Rapid access to remote memory is provided transparently
to the user with a combination of hardware and software mechanisms. This paper presents the architecture of the M-Machine
and describes how its mechanisms attempt to maximize both single thread performance and overall system throughput. The architecture
is complete and the MAP chip, which will serve as the M-Machine processing node, is currently being implemented. |
Year | DOI | Venue |
---|---|---|
1997 | 10.1007/BF02700035 | MICRO95: 28th Annual IEEE/ACM International Symposium on Microarchitecture
Ann Arbor
Michigan
USA
November, 1995 |
Keywords | Field | DocType |
multithreading,m-machine multicomputer,computer architecture,coherence.,parallelism,thread level parallelism,coherence,mesh network,functional unit,chip | Mesh networking,Multithreading,Synchronization,CPU cache,Computer science,Cache,Parallel computing,Thread (computing),Systems architecture,Throughput,Embedded system | Journal |
Volume | Issue | ISSN |
25 | 3 | 0885-7458 |
ISBN | Citations | PageRank |
0-8186-7349-4 | 85 | 14.67 |
References | Authors | |
16 | 7 |
Name | Order | Citations | PageRank |
---|---|---|---|
Marco Fillo | 1 | 106 | 18.80 |
Stephen W. Keckler | 2 | 3404 | 201.71 |
William J. Dally | 3 | 11782 | 1460.14 |
Nicholas P. Carter | 4 | 349 | 33.84 |
Andrew Chang | 5 | 85 | 14.67 |
Yevgeny Gurevich | 6 | 85 | 14.67 |
Whay Sing Lee | 7 | 99 | 15.98 |