Bare-metal and cloud infrastructure company Limestone Networks has partnered with GPU cloud platform Charg to deploy a 60 petaflops supercomputer.

Set to support AI and next-generation enterprise workloads, the system’s network backbone will be designed, built, and run by Limestone, with Charg providing the GPUs.

No further information about the GPUs Charg will deploy has been disclosed, but in a press release, the companies said the supercomputer would feature 200Gbps HDR InfiniBand fabric, suggesting the hardware will be from Nvidia. The system will also comprise an all-flash storage layer and an upcycled Cray supercomputing chassis, engineered for scalability.

The supercomputer is expected to go live in early October. Neither company has provided details about where the system will be housed.

“Timing is everything. The time is now. Limestone Networks and Charg is delivering one of the largest GPU clusters at scale,” said Preston Gosdin, CEO of Limestone Networks. “Charg brought the GPUs, we brought the global backbone, and together we built a monster. Enterprises, startups, researchers: if you’re ready to push AI past the limit, LSNMetal-Supercharg’d is open for business.”

Judd Lindvall, chief cloud officer at Charg, added: “Limestone’s ability to execute complex builds at scale made them the obvious choice. Together, we’re not just building a cloud, we’re redefining high-performance computing.”