Linux's Proposed Cache Aware Scheduling Benchmarks Show Big Potential On AMD EPYC Turin

Written by Michael Larabel in Software on 23 October 2025 at 01:00 PM EDT. Page 2 of 4. 39 Comments.

First up was running Microsoft's Ethr network measurement tool for evaluating the network performance for the localhost while involving multiple CPU threads. In coincidentally being the first benchmark fired up for this cache aware scheduling comparison, the results quickly proved jaw-dropping.

👁 Ethr benchmark with settings of Server Address: localhost, Protocol: UDP, Test: Bandwidth, Threads: 2. Cache Aware Scheduling was the fastest.

👁 Ethr benchmark with settings of Server Address: localhost, Protocol: UDP, Test: Bandwidth, Threads: 4. Cache Aware Scheduling was the fastest.

👁 Ethr benchmark with settings of Server Address: localhost, Protocol: UDP, Test: Bandwidth, Threads: 8. Cache Aware Scheduling was the fastest.

For both the TCP and UDP tests within Microsoft Ethr, there was significant advantage shown from running the Linux 6.17 kernel with the Cache Aware Scheduling patches applied.

👁 DaCapo Benchmark benchmark with settings of Java Test: Spring Boot. Cache Aware Scheduling was the fastest.

👁 DaCapo Benchmark benchmark with settings of Java Test: Apache Tomcat. Cache Aware Scheduling was the fastest.

👁 DaCapo Benchmark benchmark with settings of Java Test: Apache Xalan XSLT. Cache Aware Scheduling was the fastest.

In some of the OpenJDK Java workloads there were performance benefits too from the Cache Aware Scheduling but they tended to be mixed depending upon the particular workload.

39 Comments - Next Page

Page: 1 2 3 4 Next Page

URL: https://www.phoronix.com/review/cache-aware-scheduling-amd-turin/2

⇱ Linux's Proposed Cache Aware Scheduling Benchmarks Show Big Potential On AMD EPYC Turin Review - Phoronix

Linux's Proposed Cache Aware Scheduling Benchmarks Show Big Potential On AMD EPYC Turin