English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Modern SoCs integrate multiple CPU cores and Hardware Accelerators (HWAs)that share the same main memory system, causing interference among memoryrequests from different agents. The result of this interference, if notcontrolled well, is missed deadlines for HWAs and low CPU performance.State-of-the-art mechanisms designed for CPU-GPU systems strive to meet atarget frame rate for GPUs by prioritizing the GPU close to the time when ithas to complete a frame. We observe two major problems when such an approach isadapted to a heterogeneous CPU-HWA system. First, HWAs miss deadlines becausethey are prioritized only close to their deadlines. Second, such an approachdoes not consider the diverse memory access characteristics of differentapplications running on CPUs and HWAs, leading to low performance forlatency-sensitive CPU applications and deadline misses for some HWAs, includingGPUs.  In this paper, we propose a Simple Quality of service Aware memory Schedulerfor Heterogeneous systems (SQUASH), that overcomes these problems using threekey ideas, with the goal of meeting deadlines of HWAs while providing high CPUperformance. First, SQUASH prioritizes a HWA when it is not on track to meetits deadline any time during a deadline period. Second, SQUASH prioritizes HWAsover memory-intensive CPU applications based on the observation that theperformance of memory-intensive applications is not sensitive to memorylatency. Third, SQUASH treats short-deadline HWAs differently as they are morelikely to miss their deadlines and schedules their requests based on worst-casememory access time estimates.  Extensive evaluations across a wide variety of different workloads andsystems show that SQUASH achieves significantly better CPU performance than thebest previous scheduler while always meeting the deadlines for all HWAs,including GPUs, thereby largely improving frame rates.

SQUASH: Simple QoS-Aware High-Performance Memory Scheduler for Heterogeneous Systems with Hardware Accelerators