San Jose, CA — April 22, 2026. Pebble has achieved status as an official Google Cloud Advantage Partner. The partnership brings Pebble's autonomous energy hypervisor — software that cuts the power bill of GPU clusters and packs more compute into the same megawatts — to Google Cloud customers running AI training and inference workloads.
The partnership arrives as power becomes the binding constraint on AI. Data centers are projected to consume 6.7–12% of U.S. electricity by 2028, utility interconnect queues stretch for years, and most GPU clusters run at hardware-default power caps that strand 20–30% of usable capacity as heat. For operators, power is now both the largest controllable cost and the hardest input to scale.
Pebble's platform profiles every workload in real time and caps each GPU at its efficiency knee — the operating point where tokens-per-watt peaks. The result is a lower power bill with no SLA impact and no application changes, plus reclaimed headroom that operators can use to deploy more GPUs inside their existing power envelope.
The collaboration lets joint customers run Pebble alongside their existing Google Cloud GPU deployments to reduce demand charges and improve price-performance, without rewriting their serving stack.
About Pebble
Pebble is the autonomous energy layer for AI datacenters. By capping every GPU at its efficiency knee and turning the reclaimed power into grid-flexible capacity, Pebble helps operators cut their power bill and fit more revenue-generating compute into the same megawatts — validated live on AMD Instinct™ MI350X.
Media contact
MyPebble, Inc.
333 West San Carlos Street, Riverpark Tower, 4th Floor
San Jose, CA 95110
1-888-314-1019 · tech@gopebble.com