AI-Powered / Machine Learning Linux Performance Tuning Is Now A Thing

A year and a half ago I wrote about a start-up working on dynamically-tuned, self-optimizing Linux servers. That company is now known as Concertio and they just launched their "AI powered" toolkit for IT administrators and performance engineers to optimize their server performance.

Concertio Optimizer Studio is their product making use of machine learning that aims to optimize Linux systems with Intel CPUs for peak performance by scoping out the impact of hundreds of different tunables for trying to deliver an optimal configuration package for that workload on that hardware.

Here's how they say their Optimizer Studio software works:

Patent-pending Concertio Optimizer Studio is an easy-to-use software tool that uses AI techniques to dynamically discover near-optimal configuration settings for peak performance. Optimizer Studio software uses a workload classification engine to detect different phases of execution of a workload while system applications run. Optimizer Studio then uses reinforcement learning to evaluate and optimize each detected program phase, testing various system configuration settings, and choosing settings that maximize performance. Eventually, Optimizer Studio generates a report detailing the optimal configuration discovered to be the most beneficial for peak system performance. Performance engineers and IT professionals can then apply the discovered settings in their environment.

Unfortunately, it's not open-source. Those interested in learning more can do so at Concertio.com. If I am able to get some access to Optimizer Studio, of course, I will deliver some interesting benchmarks. From this earlier article they were using Apache in the Phoronix Test Suite and showed promising results with their earlier beta software.