BROOMFIELD, Colo.--(BUSINESS WIRE)--Liqid, the global leader in composable disaggregated infrastructure (CDI) solutions, proudly announces the successful delivery and acceptance of the largest composable supercomputers by the U.S. Department of Defense (DoD). Liqid served as the prime contractor for the $32 million contract, delivering two systems to the Army Research Laboratory DoD Supercomputing Resource Center (ARL DSRC). These systems boast the largest composable GPU pools worldwide, with 360 NVIDIA A100 GPUs. This unprecedented capability enables the ARL DSRC to meet their high-performance computing needs with exceptional speed and flexibility, seamlessly provisioning, scaling, and reallocating GPUs on-demand via software. The operations will be overseen by the High Performance Computing Modernization Program (HPCMP), which provides HPC resources, expertise, and infrastructure to support the scientific, engineering, and computational needs of the DoD.
The ARL DSRC is an advanced facility dedicated to supercomputing and computational science, serving a diverse and extensive user base within the DoD research, development, test, and evaluation (RDT&E) communities. Located at Aberdeen Proving Ground, Maryland, it operates under the Combat Capabilities Development Command (CCDC), U.S. Army Research Laboratory, Computational and Information Sciences Directorate (CISD). The CCDC ARL DSRC's primary mission is to provide world-class high-performance computing, advanced networking, and computational science tools, along with unparalleled expertise, to support the RDT&E efforts of the DoD.
The two Liqid composable supercomputers at the ARL DSRC are named "Jean" and "Kay" in honor of Jean Bartik and Kay McNulty, esteemed programmers of the pioneering ENIAC, the first programmable general-purpose electronic digital computer in the U.S.
ARL DSRC Composable HPC Technical Specifications:
Jean:
- Liqid Matrix CDI Platform
- 588 Intel Xeon Platinum 9200 Nodes
- 264 NVIDIA A100 GPUs
- 313 TB of memory
- 15.6 PB of all-flash NVMe-oF parallel file system storage
- HDR 200Gb/s InfiniBand Fabric
Kay:
- Liqid Matrix CDI Platform
- 488 Intel Xeon Platinum 9200 Nodes
- 96 NVIDIA A100 GPUs
- 226 TB of memory
- 12.9 PB of all-flash NVMe-oF parallel file system storage
- HDR 200Gb/s InfiniBand Fabric
"Modern AI/ML workloads demand massive GPU performance. By choosing Liqid, ARL can optimize the use of these powerful and invaluable resources to deliver results with utmost efficiency," said Sumit Puri, CEO & Co-founder of Liqid.
Liqid's Matrix composable platform empowers users to effortlessly manage, scale, and configure physical, bare-metal HPC systems within seconds, allowing for seamless reallocation of core data center devices based on evolving workflows and business requirements. Liqid Matrix enables dynamic rightsizing of IT resources on the fly, ensuring optimal utilization.
"We are thrilled to power the U.S. DoD with the world's largest composable supercomputers. Our unwavering commitment is to assist our customers in tackling their most pressing computing challenges with unmatched agility and speed," added Puri.
About Liqid
Liqid is the global leader in composable disaggregated infrastructure (CDI) solutions, revolutionizing the way organizations utilize IT resources. By empowering organizations with unprecedented agility, flexibility, and efficiency, Liqid enables users to effortlessly configure, scale, and manage their IT infrastructure in seconds. With Liqid’s innovative Matrix software at its core, customers get comprehensive control over resource utilization, including GPU and storage, allowing them to efficiently adapt to evolving needs. Committed to driving innovation and helping customers tackle their computing challenges, Liqid is at the forefront of shaping the future of data center infrastructure. For more information, visit www.liqid.com and follow Liqid on Twitter and LinkedIn.
