Chris Malone, who ran data center operations at OpenAI for roughly a year, has joined Nvidia as vice president of its DSX Platform. The appointment, effective September 24, 2026, represents one of the more significant talent acquisitions in the AI infrastructure space this year.
Malone originally joined OpenAI in March 2025, shortly after the Stargate announcement that brought together Oracle and SoftBank in a massive infrastructure partnership. He departed just before August 25, 2026, as part of a broader leadership reorganization within OpenAI’s infrastructure division. His responsibilities were split among several leaders, including Sachin Katti, who took on oversight of specific data center functions.
Malone’s departure coincided with a wider executive exodus that included Brad Lightcap and Fidji Simo, reflecting a strategic pivot at OpenAI away from building proprietary infrastructure and toward relying more heavily on partnerships.
What Nvidia’s DSX platform actually does
The DSX platform launched on May 31, 2026, during GTC Taipei. At its core, DSX is a reference design ecosystem, a comprehensive set of specifications and optimized software that tells data center builders exactly how to configure hardware for maximum AI performance.
The platform combines open-source software with hardware optimization, and its headline metric is tokens per watt. One standout feature is DSX MaxLPS, which uses advanced cooling techniques to allow up to 40% more GPUs to operate at peak efficiency within a given power envelope. Lambda, an early adopter, has reported up to a 24% increase in token throughput in benchmarks using the platform.
Major hardware manufacturers including Dell, HPE, Lenovo, and Supermicro are already building DSX-compatible systems.
Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.

1 hour ago
24







English (US) ·