Abstract
To reduce the miss rate of the instruction cache, a hardware-assisted loop optimization method is proposed in this paper. This method utilizes the hardware/software co-design strategy on the behavior level. Especially, this method is equipped with the specific instruction set to limit the cache misses, which can be viewed as a set of hardware for special purposes. Then based on the specific instruction set, a scheduling process is integrated which reduces the cache miss rate through the code transformation. Finally, a set of benchmarks from MediaBench1.0 are tested on the SimpleScalar platform to assist the proposed method. The final experiments indicate that 26% enhancement can be obtained for the cache miss reduction, where the specific instruction generation and the scheduling processes contribute about 23% and 3% respectively.
Original language | English |
---|---|
Title of host publication | Proceedings of the 2008 12th International Conference on Computer Supported Cooperative Work in Design, CSCWD |
Pages | 129-134 |
Number of pages | 6 |
Volume | 1 |
DOIs | |
Publication status | Published - 2008 |
Event | 2008 12th International Conference on Computer Supported Cooperative Work in Design, CSCWD - Xi'an Duration: 2008 Apr 16 → 2008 Apr 18 |
Other
Other | 2008 12th International Conference on Computer Supported Cooperative Work in Design, CSCWD |
---|---|
City | Xi'an |
Period | 08/4/16 → 08/4/18 |
Keywords
- ASIP
- Cache miss
- CSCW design
- Hardware/software co-design
ASJC Scopus subject areas
- Artificial Intelligence
- Computer Networks and Communications
- Human-Computer Interaction