LIQID Inc., one of the world’s leading software companies delivering data center composability, announced today a new case study with Durham University in Durham, England, describing its deployment of Liqid composable disaggregated infrastructure (CDI) software. Durham University is home to the COSmology MAchine (COSMA) operated by the Institute for Computational Cosmology (ICC) and is a national supercomputing facility that is part of the UK’s Distributed Research utilizing Advanced Computing (DiRAC) system. Liqid Matrix™ CDI software is part of a system used to study the origins of the universe, composing disaggregated GPUs in a tight footprint designed for a more sustainable digital architecture. With the ability to grow as resources are needed, the Liqid system will help researchers in the UK and around the world unlock the mysteries of the 14 billion-year history of the cosmos with far greater speed and efficiency than traditional data center systems.
“We are honored to work with institutions like Durham University and COSMA researchers to provide the resources to better explore the most profound questions in science,” said Sumit Puri, CEO & Cofounder, Liqid. “By providing the kind of software-defined data performance and architectural flexibility for powerful GPU, Liqid is enabling research breakthroughs that would have been impossible with traditional, static data center architectures so scientists can focus on results instead of waiting for resources and performance to become available.”
“Durham University is using cutting-edge CDI to accelerate research, improve resource utilization, and reduce the university’s carbon footprint,” said Alistair Basden, technical manager for the DiRAC Memory Intensive Service at Durham University, who was interviewed for the case study entitled Durham University’s Institute for Computational Cosmology Accelerates Results with Composability from Liqid.
The COSMA memory-intensive system is designed specifically to support the largest cosmological simulations, most notably running simulations starting with the Big Bang and propagating through the entire history of the universe. Each simulation of dark matter, dark energy, black holes, galaxies and other structures in the universe often takes months to run, followed by long periods of data analysis.
Durham University IT is also deploying Liqid Matrix-based systems as an element of its overall strategy for a more sustainable IT ecosystem. Software-based composability enables users to do more with less, increasing efficiency and curtailing the need for physical space to store the hardware while providing significant reductions in cooling and water requirements.
“Rather than populating all our servers with GPUs, we can compose the resources we need to each server. That reduces our carbon footprint,” said Basden.
Full story here.