Cyxtera, a global provider of data center colocation and interconnection services operating a footprint of 62 data centers in 29 markets around the world, has launched its Artificial Intelligence/Machine Learning (AI/ML) compute as a service offering. It would deliver data center platform capabilities enabling AI-powered workloads.
Leveraging the NVIDIA DGX A100 system, Cyxtera thinks this new offering will provide enterprises greater agility and faster deployment with a flexible OPEX model. It would avoid the need for large capital outlays and over provisioning.
Cyxtera’s solution powered by NVIDIA DGX A100 is intended to offer the simplicity and ease of the cloud, with the deterministic performance of dedicated infrastructure. This new, flexible infrastructure model leverages “point-and-click” provisioning via the company’s data center services exchange.
Cyxtera’s colocation clients will also have direct access to a rich ecosystem of service providers and technology solutions. This includes storage/Storage as a Service (StaaS), interconnectivity, and security, among other managed services, to complement their NVIDIA DGX A100 deployment via the Cyxtera Marketplace.
“Launching our DGX-based AI/ML compute as a service offering reinforces our continued focus on providing enterprise customers with a rich variety of leading-edge infrastructure options across our global footprint,” said Russell Cozart, Senior Vice President of Marketing and Product Strategy for Cyxtera. “As a certified NVIDIA DGX-Ready Program partner, we’re offering accelerated solutions ideally suited to our enterprise customer base.”
Managed Service Providers
Enterprises will be able to provision NVIDIA DGX A100 systems via Cyxtera’s service exchange. This would provide the security and control of a single-tenant, dedicated infrastructure combined with the flexibility and agility of cloud. The NVIDIA DGX A100 offering also allows managed service providers (MSPs) to focus on service delivery by eliminating large, upfront capital investments in data center infrastructure while providing access to customers via the Cyxtera exchange.
“Many organizations want to speed up their AI innovation cycle, but lack the infrastructure to support it,” said Charlie Boyle, Vice President, General Manager of DGX Systems at NVIDIA. “Cyxtera’s AI/ML compute-as-a-service offering makes it easy to leverage the performance, flexibility and efficiency of DGX A100 in a cost-effective model, enabling more businesses to succeed in the race to deploy AI.”
The DGX-based AI/ML compute as a service offering will initially be available in Cyxtera’s Northern Virginia, Dallas/Fort Worth, and London metro data centers, with plans to expand across the company’s global footprint of data centers.