论文信息 - Sensitivity Analysis of Core Specialization Techniques

Sensitivity Analysis of Core Specialization Techniques

The instruction footprint of OS-intensive workloads such as web servers, database servers, and file servers typically exceeds the size of the instruction cache (32 KB). Consequently, such workloads incur a lot of i-cache misses, which reduces their performance drastically. Several papers have proposed to improve the performance of such workloads using core specialization. In this scheme, tasks with different instruction footprints are executed on different cores. In this report, we study the performance of five state of the art core specialization techniques: SelectiveOffload [6], FlexSC [8], DisAggregateOS [5], SLICC [2], and SchedTask [3] for different system parameters. Our studies show that for a suite of 8 popular OS-intensive workloads, SchedTask performs best for all evaluated configurations.

Prathmesh Kallurkar | Smruti R. Sarangi

[1] Min Lee,et al. Memory region: a system abstraction for managing the complex memory structures of multicore platforms , 2013 .

[2] James E. Smith,et al. Data Cache Prefetching Using a Global History Buffer , 2005, IEEE Micro.

[3] David W. Nellans,et al. Interference Aware Cache Designs for Operating System Execution , 2009 .

[4] Anastasia Ailamaki,et al. SLICC: Self-Assembly of Instruction Cache Collectives for OLTP Workloads , 2012, 2012 45th Annual IEEE/ACM International Symposium on Microarchitecture.

[5] Prathmesh Kallurkar,et al. SchedTask: A Hardware-Assisted Task Scheduler , 2017, 2017 50th Annual IEEE/ACM International Symposium on Microarchitecture (MICRO).

[6] Jignesh M. Patel,et al. Call graph prefetching for database applications , 2003, TOCS.

[7] Michael Stumm,et al. FlexSC: Flexible System Call Scheduling with Exception-Less System Calls , 2010, OSDI.