University of Birmingham deploys UK’s largest IBM POWER9 AI cluster
FYI, this story is more than a year old
The largest IBM POWER9 Artificial Intelligence (AI) cluster in the UK is soon to be calling the University of Birmingham home.
The University initially deployed two IBM Power Systems AC922 servers that were powered by POWER9 CPUs in September this year, but the Advanced Research Computing (ARC) team soon realised that it needed more horsepower to cater for the ever-increasing AI workloads.
In light of this, the University will now be deploying an additional nine IBM Power Systems AC922 warm water-cooled nodes, each equipped with four NVIDIA Tesla V100 16GB Tensor Core GPUs, 1TB of system memory, dual 18 core POWER9 CPUs and Mellanox 100Gb EDR InfiniBand.
Running on IBM PowerAI Enterprise software, the University hopes to capitalise on the UK’s largest IBM POWER9 cluster. Working with high performance computing (HPC), storage and data analytics integrator OCF, the University will integrate it’s a total of 11 IBM servers will into its existing HPC infrastructure, the Birmingham Environment for Academic Research (BEAR).
“It’s very important to us as a research-led institution that we are at the forefront of data research which means we are always looking at ways to make AI quicker and more accessible for our researchers,” says University of Birmingham research computing infrastructure architect Simon Thompson.
“With the sheer amount of data, the common questions from researchers are how can we analyse it fast enough and how can we make the process even quicker? With our early deployment of the two IBM POWER9 servers we have seen what is possible. By scaling up, we can keep-pace with the escalating demand and offer the computational capacity and capability to attract leading researchers to the University.”
According to Thompson, the result of the substantial enhancement to BEAR will be an even more powerful and versatile computing environment to serve researchers.
For example, fellows from The Alan Turing Institute looking at early diagnosis of and new therapies for heart disease and cancer, will use AI to run faster diagnostics in the future. Meanwhile, researchers in the physical sciences are similarly using machine learning and data science approaches to quantify the 4D (3D plus time) microstructures of advanced materials collected at national large synchrotron facilities such as the Diamond Light Source.
"We are thrilled that the University of Birmingham has decided to invest in building the UK's largest POWER9 AI cluster,” says IBM Servers UK & Ireland director Simon Robertson.
“We are proud to see the practical application of IBM technology used by researchers across the University and beyond."
OCF managing director Julian Fielden says they are delighted to be working with the University on this project.
“AI workloads are driving data intensive challenges that can only be met with accelerated infrastructure, such as IBM’s POWER9. The University is leading the way with this impressive project and will continue to attract world-class researchers with this type of innovation,” says Fielden.
Energy efficiency is also a priority in the initiative, with the University investing in the reduction of energy consumption and boasting the UK’s first purpose-built water-cooled research-focused data centre. 85 percent of the heat is recovered directly through the cooling systems to minimise the cooling overgeads.
In addition, the IBM POWER9-based AC922 servers include warm water-cooled nodes, where water is taken directly across the CPUs and GPUs at temperatures up to 35C. The data centre uses no air-cooling systems and accommodates the IBM systems running alongside “direct to node” water cooled technology from other vendors.
“With OCF as our partner, we have the right guidance, skills and expertise to continually move BEAR forward,” says Thompson.
“It’s not just about having the largest cluster, it’s about providing cutting-edge HPC tools for our researchers from both traditional and non-traditional disciplines, who will be able to process data faster and generate new findings, achieving greater research impact.”