Yokogawa — leaderboard ()

800 VDC AI data centers get microsecond peak shaving via SST + storage tie-up

DG Matrix and Skeleton Technologies are integrating Skeleton’s fast-response energy storage systems with DG Matrix’s Interport solid-state transformer platform as the data center industry moves toward 800 VDC power infrastructure for AI deployments.

The work centers on DG Matrix’s galvanically isolated Interport multi-port solid-state transformer (SST) platform and spans multiple Interport form factors: the Interport CM medium-voltage SST skid, which converts 12–34.5 kV AC directly to high-voltage DC; the Interport CL “sidecar” aimed at mitigating GPU pulse loads in the white space; and Interport Flex, an outdoor grey-space product that can be skid-mounted and paralleled into an Interport 360 end-to-end solution.

Skeleton’s systems are intended to handle peak-shaving and backup power duties, with response times described as ranging from microseconds to minutes. The combined architecture targets a problem operators are increasingly designing around in AI rooms: extreme, repetitive GPU cluster pulse loads that can stress upstream infrastructure and outpace the dynamic response of conventional UPS and battery systems. DG Matrix’s Interport is described as routing power in real time across AC and DC sources, loads, and storage at multiple voltage levels, with “single-stage conversion” and software-defined, multi-port power routing.

For data center engineers, the practical takeaway is that 800 VDC doesn’t just change bus bars and protection schemes—it changes how you think about transient behavior at the rack and row level. If GPU load steps are the dominant disturbance, then the placement and response time of energy storage relative to the IT load becomes a first-order design constraint, not a nice-to-have.

Skeleton Technologies CEO and co-founder Taavi Madiberk said, “As power demands for AI workloads become even greater, fast-acting energy storage is essential to ensuring that GPU bursts don’t destabilize electric grids.” DG Matrix CEO and co-founder Haroon Inam said that integrating Skeleton’s storage with Interport is intended to “advance a more complete architecture for AI data centers: one designed for dynamic pulse loads, resilient power, and faster deployment.”

The companies also pointed to SemiAnalysis research projecting about 39 GW of incremental AI data center capacity served by 800 VDC architecture by 2030, and an SST total addressable market of about $13 billion. DG Matrix said Interport natively supports 800 VDC today, and Skeleton cited its GrapheneCBU800 and GrapheneBBU800 systems as built for the same voltage class. DG Matrix is at dgmatrix.com, and Skeleton Technologies is at skeletontech.com.

Source: DG Matrix

Yokogawa — below-posts ()

Get Data Center Engineering News In Your Inbox:

By subscribing, you agree to our Privacy Policy and Terms of Use. You can unsubscribe at any time.

Sponsored: Infineon at OCP 2026: Exclusive Content Hub. Executive interviews, technical documents, reference designs, and live expert panel. Unlock the hub.
Yokogawa — tower ()

Popular Posts:

Engineering AI Rack Power: Infineon Answers the Questions Asked at OCP, live expert panel with Carl Smith, Davide Chiola and Michael Williams
AI data center architecture evolution and key findings from OCP
thumb-dc1-selection-guide
Data center semiconductor selection guide
thumb-sst1-solid-state-transformers
Why line-frequency transformers fall short in AI data center power design
thumb-dc2-future-of-powering-ai
Redefining the power flow in and beyond AI data centers
Dr. Fanny Björk, Infineon — Microseconds, not milliseconds
Grid to core: the case for shifting to DC-grids

Share Your Data Center Engineering News

Do you have a new product announcement, webinar, whitepaper, or article topic? 

Chroma — tower

Get Data Center Engineering News In Your Inbox:

By subscribing, you agree to our Privacy Policy and Terms of Use. You can unsubscribe at any time.

Chroma — leaderboard