Utilizing shared data in chip multiprocessors with the Nahalal architecture
Proceedings of the twentieth annual symposium on Parallelism in algorithms and architectures
Three-dimensional Integrated Circuit Design
Three-dimensional Integrated Circuit Design
Cache-aware partitioning of multi-dimensional iteration spaces
SYSTOR '09 Proceedings of SYSTOR 2009: The Israeli Experimental Systems Conference
Last Bank: Dealing with Address Reuse in Non-Uniform Cache Architecture for CMPs
Euro-Par '09 Proceedings of the 15th International Euro-Par Conference on Parallel Processing
Performance and power aware CMP thread allocation modeling
HiPEAC'10 Proceedings of the 5th international conference on High Performance Embedded Architectures and Compilers
| Hi-index | 0.00 |
This paper addresses cache organization in Chip Multiprocessors (CMPs). We show that in CMP systems it is valuable to distinguish between shared data, which is accessed by multiple cores, and private data accessed by a single core. We introduce Nahalal, an architecture whose novel floorplan topology partitions cached data according to its usage (shared versus private data), and thus enables fast access to shared data for all processors while preserving the vicinity of private data to each processor. Nahalal exhibits significant improvements in cache access latency compared to a traditional cache design.