· 2012
Cache-Conscious Wavefront Scheduling
- 421citations
- 2012year
Short summary
A new hardware scheduling mechanism, Cache-Conscious Wavefront Scheduling (CCWS), reduces L1 data cache misses by an average of 25%, improving performance by 24% on cache-sensitive GPU workloads.
AI-generated from the title and abstract; the full text is not read.
Key points
- Proposes Cache-Conscious Wavefront Scheduling (CCWS), an adaptive hardware mechanism for GPUs.
- CCWS utilizes an intra-wavefront locality detector to shape memory access patterns and avoid L1 cache thrashing.
- Achieves an average 25% reduction in L1 data cache misses.
- Results in a harmonic mean 24% performance improvement over existing policies on cache-sensitive workloads.
- Estimated chip area cost is only 0.17%.
AI-generated from the title and abstract; the full text is not read.
Abstract
This paper studies the effects of hardware thread scheduling on cache management in GPUs. We propose Cache-Conscious Wave front Scheduling (CCWS), an adaptive hardware mechanism that makes use of a novel intra-wave front locality detector to capture locality that is lost by other schedulers due to excessive contention for cache capacity. In contrast to improvements in the replacement policy that can better tolerate difficult access patterns, CCWS shapes the access pattern to avoid thrashing the shared L1. We show that CCWS can outperform any replacement scheme by evaluating against the Belady-optimal policy. Our evaluation demonstrates that cache efficiency and preservation of intra-wave front locality become more important as GPU computing expands beyond use in high performance computing. At an estimated cost of 0.17% total chip area, CCWS reduces the number of threads actively issued on a core when appropriate. This leads to an average 25% fewer L1 data cache misses which results in a harmonic mean 24% performance improvement over previously proposed scheduling policies across a diverse selection of cache-sensitive workloads.
The authors' abstract, as published at the source., 2012 · DOI ↗
Continue with a free account
Ask the paper: 3 free questions a day about this paper; save it, get its citation, new summaries every day for your field. Takeaways are Premium.
Continue free on the webSign in with Google or Apple; no card needed. You come back to this paper.
On your phone:
Field: Hardware and Architecture
Hardware and ArchitectureComputer Science