Performance Evaluation of a Multicore System with Optically Connected Memory Modules
NOCS '10 Proceedings of the 2010 Fourth ACM/IEEE International Symposium on Networks-on-Chip
| Hi-index | 0.00 |
Large systems architectural ideas progressed very rapidly during the middle of the 60's but is currently in a sort of hiatus. This hiatus is largely due to the need to digest the significance of the preceding developments and is currently reinforced by the economic recession. These architectural developments have yielded announced or about to be announced products like the IBM 195, the CDC 7600 and STAR, and the ILLIAC IV. These machines represent not only different solutions to the problem of high performance, but also different views of what constitutes the problem. These concepts of problems, along with their solutions, goes from: consisting of identical processing performed at each of a great many points---so put in many execution units run by a single instruction control---and ranges to: problems require quite diverse functions rather locally---so make it possible to utilize these diverse facilities concurrently. These machines above are listed in the order of their degree of independence of problem regularity when it occurs. The common denominator of these machines is the intent to capitalize upon the existence of concurrently executable sequentially independent process strings.