Vast Data is teaming up with SK Telecom on the latter's AI cluster.

Comprised of Nvidia Blackwell GPUs and dubbed the Haein Cluster, Vast will be providing its AI operating system to enable SKT to create a fully virtualized GPU-as-a-Service (GPUaaS) environment.

The Haein Cluster was launched earlier this month with 1,000 Nvidia B200 GPUs, building on SK Telecom's existing offering based on Nvidia H100s that was launched in December 2024.

The GPUaaS is available via the Petasus AI Cloud, which has now been integrated with Vast Data's disaggregated, shared-everything (DASE) platform on Nvidia HGX server architecture, built by Supermicro. According to Vast, this enables the offering to have a high-throughput, secure, and multi-tenant environment that supports both AI training and inference at scale, while remaining sovereign to South Korea.

“Vast Data’s unified architecture has been instrumental in helping us move from legacy bare-metal deployments to a fully virtualized, production-grade AI cloud,” said DK Lee, vice president, and head of the AI DC Lab at SK Telecom. “The Vast AI OS powers the performance, simplicity, and flexibility needed to support the next generation of sovereign AI workloads, and gives us the confidence to scale fast and securely. With Vast, we’re enabling a GPUaaS platform that meets the exacting needs of government, research, and enterprise AI customers in South Korea.”

“SK Telecom is defining the future of national-scale AI infrastructure, and VastT is proud to support their vision,” added Sunil Chavan, VP, APAC, Vast Data. “From our earliest conversations, it was clear that SKT needed cutting-edge infrastructure to match the speed and complexity of enterprise-grade uptime and nation-state inference and training. By eliminating traditional bottlenecks around data movement, provisioning, and security, Vast is enabling SKT to launch a sovereign and secure AI infrastructure that offers speed and flexibility at scale for Korea."

In addition to a GPUaaS, Haein was selected by the South Korean Ministry of Science and ICT for the "Proprietary Found Model" project for the development of national AI models.

The B200 cluster is based in the company's Gasan, Seoul data center, the same as its prior H100 GPUaaS. The H100 GPUaaS was launched in partnership with Lambda, having teamed up with the AI cloud company in August 2024, with plans to deploy Lambda's platform in the SK data center.

The Gasan data center launched in 2021, with 10 stories above ground and another five basement floors. At the time, it could support 46MW of IT Load across 69,300 sqm (745,940 sq ft). SKT said that the site can currently support up to 44kW per rack.