Victim-Caching for Large Caches and Modern Workloads
نویسنده
چکیده
Jouppi 4] has proposed that direct-mapped caches should be assisted by victim caches: small, fully-associative caches which reduce connict misses by saving the blocks most recently thrown out of the main cache. Jouppi found that victim-caching worked well for 4KB data caches. Here, we continue this work, studying victim-caching for modern workloads (SPEC95) and larger caches (32{64KB). We nd that, if victim-cache size is allowed to scale up to 1 4 of main-cache size, victim-caching continues to work well at 32KB, but shows signs of breaking down for larger cache sizes. We conclude that victim-caching is a credible way of reducing miss-rates for modern, medium-sized caches, and that it may be particularly useful for reducing chip-area costs.
منابع مشابه
Selective Victim Caching: A Method to Improve the Performance of Direct-Mapped Caches
Victim caching was proposed by Jouppi [4] as an approach to improve the miss rate of direct-mapped caches. This approach augments the direct-mapped main cache with a srnall fully-associate cache, called victim cache, that stores cache blocks evicted from the main cache as a result of replacements. We propose and evaluate an improvement of this scheme, called selective victim caching. In this sc...
متن کاملDesign and Performance Evaluation of a Cache Assist to implement Selective Caching
EEcient instruction and data caches are extremely important for achieving good performance from modern high performance processors. Conventional cache architectures exploit locality, but do so rather blindly. By forcing all references through a single structure, the cache's eeectiveness on many references is reduced. This paper presents a selective caching scheme for improving cache performance...
متن کاملAnalysis of Sub-block Placement and Victim Caching Techniques
Rapid advances in computer technology have led to the development of processors with peak performances of the order of GHz. Since it is not feasible to have unlimited fast-memory, the performance of these processors is handicapped if the performance of the memory-hierarchy is poor. Caching techniques have been developed with this in mind. This paper presents the analysis of the performance of t...
متن کاملComputer Science and Artificial Intelligence Laboratory Victim Migration: Dynamically Adapting Between Private and Shared CMP Caches
Future CMPs will have more cores and greater onchip cache capacity. The on-chip cache can either be divided into separate private L2 caches for each core, or treated as a large shared L2 cache. Private caches provide low hit latency but low capacity, while shared caches have higher hit latencies but greater capacity. Victim replication was previously introduced as a way of reducing the average ...
متن کاملVictim Migration: Dynamically Adapting Between Private and Shared CMP Caches
Future CMPs will have more cores and greater onchip cache capacity. The on-chip cache can either be divided into separate private L2 caches for each core, or treated as a large shared L2 cache. Private caches provide low hit latency but low capacity, while shared caches have higher hit latencies but greater capacity. Victim replication was previously introduced as a way of reducing the average ...
متن کامل