Scalable directoryless shared memory coherence using execution migration
We introduce the concept of deadlock-free migration-based coherent shared memory to the NUCA family of architectures. Migration-based architectures move threads among cores to guarantee sequential semantics in large multicores. Using a execution migration (EM) architecture, we achieve performance co...
| Main Authors: | , , , , |
|---|---|
| Other Authors: | |
| Published: |
2010
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/1721.1/60039 |
| _version_ | 1811086511995617280 |
|---|---|
| author | Lis, Mieszko Shim, Keun Sup Cho, Myong Hyon Khan, Omer Devadas, Srinivas |
| author2 | Srini Devadas |
| author_facet | Srini Devadas Lis, Mieszko Shim, Keun Sup Cho, Myong Hyon Khan, Omer Devadas, Srinivas |
| author_sort | Lis, Mieszko |
| collection | MIT |
| description | We introduce the concept of deadlock-free migration-based coherent shared memory to the NUCA family of architectures. Migration-based architectures move threads among cores to guarantee sequential semantics in large multicores. Using a execution migration (EM) architecture, we achieve performance comparable to directory-based architectures without using directories: avoiding automatic data replication significantly reduces cache miss rates, while a fast network-level thread migration scheme takes advantage of shared data locality to reduce remote cache accesses that limit traditional NUCA performance. EM area and energy consumption are very competitive, and, on the average, it outperforms a directory-based MOESI baseline by 6.8% and a traditional S-NUCA design by 9.2%. We argue that with EM scaling performance has much lower cost and design complexity than in directory-based coherence and traditional NUCA architectures: by merely scaling network bandwidth from 128 to 256 (512) bit flits, the performance of our architecture improves by an additional 8% (12%), while the baselines show negligible improvement. |
| first_indexed | 2024-09-23T13:27:06Z |
| id | mit-1721.1/60039 |
| institution | Massachusetts Institute of Technology |
| last_indexed | 2024-09-23T13:27:06Z |
| publishDate | 2010 |
| record_format | dspace |
| spelling | mit-1721.1/600392019-04-10T15:44:51Z Scalable directoryless shared memory coherence using execution migration Lis, Mieszko Shim, Keun Sup Cho, Myong Hyon Khan, Omer Devadas, Srinivas Srini Devadas Computation Structures multicore memory architecture cache coherence nuca We introduce the concept of deadlock-free migration-based coherent shared memory to the NUCA family of architectures. Migration-based architectures move threads among cores to guarantee sequential semantics in large multicores. Using a execution migration (EM) architecture, we achieve performance comparable to directory-based architectures without using directories: avoiding automatic data replication significantly reduces cache miss rates, while a fast network-level thread migration scheme takes advantage of shared data locality to reduce remote cache accesses that limit traditional NUCA performance. EM area and energy consumption are very competitive, and, on the average, it outperforms a directory-based MOESI baseline by 6.8% and a traditional S-NUCA design by 9.2%. We argue that with EM scaling performance has much lower cost and design complexity than in directory-based coherence and traditional NUCA architectures: by merely scaling network bandwidth from 128 to 256 (512) bit flits, the performance of our architecture improves by an additional 8% (12%), while the baselines show negligible improvement. 2010-11-23T20:30:14Z 2010-11-23T20:30:14Z 2010-11-22 http://hdl.handle.net/1721.1/60039 MIT-CSAIL-TR-2010-053 18 p. application/pdf |
| spellingShingle | multicore memory architecture cache coherence nuca Lis, Mieszko Shim, Keun Sup Cho, Myong Hyon Khan, Omer Devadas, Srinivas Scalable directoryless shared memory coherence using execution migration |
| title | Scalable directoryless shared memory coherence using execution migration |
| title_full | Scalable directoryless shared memory coherence using execution migration |
| title_fullStr | Scalable directoryless shared memory coherence using execution migration |
| title_full_unstemmed | Scalable directoryless shared memory coherence using execution migration |
| title_short | Scalable directoryless shared memory coherence using execution migration |
| title_sort | scalable directoryless shared memory coherence using execution migration |
| topic | multicore memory architecture cache coherence nuca |
| url | http://hdl.handle.net/1721.1/60039 |
| work_keys_str_mv | AT lismieszko scalabledirectorylesssharedmemorycoherenceusingexecutionmigration AT shimkeunsup scalabledirectorylesssharedmemorycoherenceusingexecutionmigration AT chomyonghyon scalabledirectorylesssharedmemorycoherenceusingexecutionmigration AT khanomer scalabledirectorylesssharedmemorycoherenceusingexecutionmigration AT devadassrinivas scalabledirectorylesssharedmemorycoherenceusingexecutionmigration |