CAPS: Cache Allocation with Partial Sharing
نویسندگان
چکیده
In a multicore system, effective management of shared last level cache (LLC) has attracted significant research attention. However, almost none of the existing solutions had been implemented on a real system until Intel introduced Cache Allocation Technology (CAT) to its commodity processors recently. CAT itself implements way partitioning and thus can only allocate at a coarse granularity, which does not scale well for a large thread or program count to serve their various performance goals effectively. We overcome these limitations by deliberately and precisely sharing part of the allocations among programs and cores.
منابع مشابه
Optimal risk allocation for regulated monopolies and consumers
The model shows how a regulated monopolist’s price should change as random cost and demand parameters are revealed. The regulator has a Ramsey-type problem. With a linear tariff a trade-off between allocative efficiency and risk sharing typically exists. The attitudes of the consumer and the firm to both income and price risk determine how the price should move. Sufficient conditions are found ...
متن کاملCAPS: A peer data sharing system for load mitigation in cellular data networks
The exponential growth of mobile data users and services places a heavy burden on the limited wireless bandwidth of cellular data networks. The situation will be exacerbated with the advent of high bandwidth multimedia applications for mobile devices. In this paper, we propose an architecture called the Cellular-based Ad hoc Peer Data Sharing system (CAPS) to reduce the load on the cellular net...
متن کاملStudy of the Effects of Program Inputs on Co-Scheduling
Cache sharing in Chip Multiprocessors (CMPs) helps hide the latency of inter-thread communications and improve effective memory bandwidth, but also brings cache contention issues, which often cause cache thrashing and drastic performance degradation. Recent studies show the effectiveness of job co-scheduling in alleviating cache contention, but without exploring the impact of program inputs on ...
متن کاملPhotonic Architectures for Distributed Shared Memory Multiprocessors
This paper studies the interaction between the access protocol used to provide arbitration for a wavelengthdivision multiple access photonic network and the cache coherence protocol required to support a distributed shared memory environment. The architecture is based on wavelength division multiplexing which enables multiple multi-access channels to be realized on a single optical fiber. Large...
متن کاملCaPPS: cache partitioning with partial sharing for multi-core embedded systems
As the number of cores in chip multi-processor systems increases, the contention over shared last-level cache (LLC) resources increases, thus making LLC optimization critical, especially for embedded systems with strict area/energy/power constraints. We propose cache partitioning with partial sharing (CaPPS), which reduces LLC contention using cache partitioning and improves utilization with sh...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2017