– NVIDIA DGX-Ready status enables Kao Data to support and host the latest NVIDIA accelerated computing technology
– Provides a dedicated technical environment for start-ups, scale-ups and world-leading organisations utilising GPU-accelerated computing on an industrial-scale
– Architectural requirements of its OCP-Ready™ facility provides seamless integration with NVIDIA DGX systems, and establishes a home for HPC and AI within the UK Innovation Corridor
Kao Data, the specialist developer and operator of advanced, carrier neutral data centres for high performance colocation, has announced it has become NVIDIA DGX-Ready Data Center certified. This step further reinforces its capabilities as a world-class data centre operator and highlights its expertise in delivering customised architectures for high-performance computing (HPC), AI and supercomputing environments.
Recommended ITech News: Viking Enterprise Solutions Introduces New NVMe-over-Fabrics Enterprise Storage Solution
The DGX-Ready Data Center status officially certifies Kao Data’s capability to support and host the latest breakthroughs in NVIDIA GPUs such as the NVIDIA A100 Tensor Core GPU, an area in which the organisation is seeing increased adoption through its work with customers in Cambridge and the UK Innovation Corridor. This also includes NVIDIA’s DGX SuperPOD, which provides the digital foundations for comprehensive advances in machine learning and deep learning technologies.
“DGX-Ready Data Center architectures are quickly changing both the dynamics and capabilities of data centres built for HPC, AI and supercomputing,” said Gérard Thibault, CTO, Kao Data. “From inception, Kao Data was precision engineered to deliver a dedicated, robust and resilient environment for intensive computing. Whether a solo DGX system, or something far larger like the acclaimed NVIDIA Cambridge-1 supercomputer, Kao Data is the perfect home for these forms of advanced computing.”
Recommended ITech News: Sandvine Launches Industry’s First Service Innovation and Intelligence Portfolio for 5G, Cloud and Edge Networks
Designed to be ultra-energy efficient, Kao Data offers a design Power Utilisation Effectiveness (PUE) of 1.2 and provides data halls with concrete slab-flooring, wide containment aisles, and no columns. Kao Data allows data halls to be customised to suit air-cooled HPC architectures with various rack power densities. NVIDIA customised Kao Data’s power and cooling infrastructure to easily accommodate a DGX A100 SuperPOD deployment, such as the NVIDIA Cambridge-1, which is expected to rank among the world’s top 5 most energy-efficient supercomputers on the Green500 list, being both efficiently cooled and powered by renewable energy.
“Cambridge and the UK Innovation Corridor is a transformative technology hub, and a significant growth region for HPC and AI,” said Marc Hamilton vice president of solutions architecture and engineering at NVIDIA. “Having a DGX-Ready Data Center partner such as Kao Data within this digital ecosystem is a crucial step that reinforces our customer commitments and supports them throughout the journey.”
Recommended ITech News: WANdisco Deepens Product Integration with Databricks to Accelerate Time to Value for Cloud-Scale Analytics
