Mitigating Multi-bit Soft Errors in L1 Caches Using Last-Store Prediction
نویسندگان
چکیده
Recent studies suggest that the rate of spatial multi-bit soft errors will increase with future technology scaling. Unfortunately, multi-bit errors cannot be effectively mitigated with conventional techniques in L1 data caches (e.g., bit interleaving or stronger coding) due to high power and/or latency overheads. We propose the laststore predictor, a lightweight prediction mechanism that accurately determines when a cache block is written for the last time and writes the data back to the L2 cache where increased access latency permits more effective multi-bit error protection. Using a combination of commercial workloads and SPEC CPU2000 benchmarks, we show that, on average, write-back L1 data caches are 42% vulnerable to multi-bit soft errors. Where SECDED ECC fails to mitigate multi-bit errors, our mechanism reduces the multi-bit soft-error vulnerability to 12% on average.
منابع مشابه
Evaluating Application Vulnerability to Soft Errors in Multi-level Cache Hierarchy
As the capacity of caches increases dramatically with new processors, soft errors originating in cache memories has become a major reliability concern for high performance processors. This paper presents application specific soft error vulnerability analysis in order to understand an application’s responses to soft errors from different levels of caches. Based on a high-performance processor si...
متن کاملR2Cache: Reliability-aware reconfigurable last-level cache architecture for multi-cores
On-chip last-level caches in multicore systems are one of the most vulnerable components to soft errors. However, vulnerability to soft errors highly depends upon the parameters and configuration of the last-level cache, especially when executing different applications. Therefore, in a reconfigurable cache architecture, the cache parameters can be adapted at run-time to improve its reliability ...
متن کاملSoft-Error Tolerant Cache Architectures
The problem of soft errors caused by radiation events are expected to get worse with technology scaling. This thesis focuses on mitigation of soft errors to improve the reliability of memory caches. We survey existing mitigation techniques and discuss their issues. We then propose 1) a technique that can mitigate soft errors in caches with lower costs than the widely-used Error Correcting Code ...
متن کاملCache Compression through Noise Prediction
Caches are very inefficiently utilized because not all the excess data fetched into the cache, to exploit spatial locality, is utilized. Our experiments show that Level 1 data cache has a utilization (in terms of the percentage of data brought into the cache that is actually used) of only about 45%. Our studies also show that a prediction accuracy of about 95% can be achieved when predicting th...
متن کامل2.3 Cache memory management for performance and reliability
The placement of the Last Level Cache (LLC) banks in the GPU on-chip network can significantly affect the performance of memory-intensive workloads. In this paper, we attempt to offer a placement methodology for the LLC banks to maximize the performance of the on-chip network connecting the LLC banks to the streaming multiprocessors in GPUs. We argue that an efficient placement needs to be deri...
متن کامل