Techniques for reducing power consumption in CMP NUCA caches
نویسندگان
چکیده
Current trend of technology scaling makes it possible to put a huge number of transistors on a single die. While dynamic power consumption can benefit from technology scaling, static power consumption get worse, thus making the latter the dominant factor of power consumption in future microprocessor systems. As on-chip cache memories require the most part of chip area and number of transistors, techniques have been developed to improve their power efficiency. In this paper, we analyze, in the context of CMP systems, how applications use cache banks depending on their locality, in order to motivate the opportunity of applying power-saving techniques to large L2 cache in a CMP environment.
منابع مشابه
Techniques for reducing power consumption in CMP Nuca cache
Current trend of technology scaling makes it possible to put a huge number of transistors on a single die. While dynamic power consumption can benefit from technology scaling, static power consumption get worse, thus making the latter the dominant factor of power consumption in future microprocessors system. As on-chip cache memories require the most part of chip area and number of transistors,...
متن کاملWay adaptable D-NUCA caches
Non-uniform cache architecture (NUCA) aims to limit the wire-delay problem typical of large on-chip last level caches: by partitioning a large cache into several banks, with the latency of each one depending on its physical location and by employing a scalable on-chip network to interconnect the banks with the cache controller, the average access latency can be reduced with respect to a traditi...
متن کاملEvaluating Power Consumption of D-NUCA caches
D-NUCA caches are on-chip cache memories characterized by multi-bank partitioning and data migration. They exhibit high hit rates while keeping the access latency low. As counterpart, such caches are affected by high static and dynamic power consumption. In this work we present a preliminary power consumption evaluation of a D-NUCA cache. Results show the existing balance among static and dynam...
متن کاملShrinking L1 Instruction Caches to Improve Energy-Delay in SMT Embedded Processors
Instruction caches are responsible for a high percentage of the chip energy consumption, becoming a critical issue for battery-powered embedded devices. We can potentially reduce the energy consumption of the first level instruction cache (L1-I) by decreasing its size and associativity. However, demanding applications may suffer a dramatic performance degradation, specially in superscalar multi...
متن کاملJigsaw: Scalable Software-Defined Caches (Extended Version)
Shared last-level caches, widely used in chip-multiprocessors (CMPs), face two fundamental limitations. First, the latency and energy of shared caches degrade as the system scales up. Second, when multiple workloads share the CMP, they suffer from interference in shared cache accesses. Unfortunately, prior research addressing one issue either ignores or worsens the other: NUCA techniques reduce...
متن کامل