May 30, 2022 — OCF, the high-performance compute (HPC), storage, cloud and AI integrator, has partnered with Energy Aware Solutions (EAS) to provide customers with the best energy optimization for HPC and AI. The partnership reinforces OCF’s commitment to sustainability and helps towards the reduction of power consumption used by HPC clusters and datacentres.
HPC installations are required to provide maximum results from the power they consume. With the integration of EAS’s Energy Aware Runtime (EAR) open-source software, job power and performance metrics are automatically measured and node settings are optimized to run as efficiently as possible, saving around 10 percent of energy consumption, without any user intervention.
EAR provides power and energy control, accounting, monitoring and optimization of applications running on the cluster and the overall global cluster in a fully transparent way to the users. By increasing data center energy optimization, there is a reduction in energy costs and the assurance of eco-responsibility for organizations using HPC and AI.
Mischa van Kesteren, Sustainability Officer at OCF says: “We are excited to be offering EAS’s energy optimization solution to our customers. We are committed to sustainability and making a larger impact where we can and this partnership gives us the chance to provide a new service for HPC and AI users wanting to monitor and improve their energy consumption in a more responsible and cost-effective way.
The research performed by our customers is essential to combating many of the challenges we face, be they medical, mechanical or environmental. With EAS, we can enable them to maximize their research per watt to solve these challenges with the lowest energy impact possible.”
Luigi Brochard, CEO at Energy Aware Solutions says: “The main focus of EAS is to reduce energy consumption, it’s simple. We are delighted to be partnering with OCF, to give customers the opportunity to look at more sustainable ways of maintaining their HPC and AI systems and reducing their energy bills. EAR is a powerful tool that is completely transparent and can make changes in real time to energy consumption without the need of intervention.”
HPC and AI are highly resource intensive workloads for IT hardware with networking, memory, storage and compute resources often being pushed to their maximum. However, few applications stress all elements simultaneously. By enabling EAS, HPC and AI environments will profile their applications and then reduce the power consumption of individual components (memory DIMMS, CPUs, GPUs, etc…)
, during a job run. This is how it can reduce energy consumption in a transparent and non-disruptive manner.