Composable infrastructure company GigaIO has announced the general availability of its portable AI supercomputer, Gryf.

Co-designed by GigaIO and SourceCode, Gryf, which is described by the companies as “the world’s first suitcase-sized AI supercomputer,” delivers data center-scale compute power directly to Edge operations, enabling real-time intelligence and analytics.

GigaIO Gryf
GigaIO's Gryf AI supercomputer – GigaIO

In a statement, GigaIO said the machine is designed and built in the US, with the company having already secured “significant orders” from the US Department of Defense and the intelligence community. Gryf can also support applications across healthcare, scientific research, industry and manufacturing, oil and gas, and sports analytics.

Powered by the company’s FabreX AI memory fabric, GigaIO said Gryf provides “unprecedented computing power in a ruggedized, field-ready design,” meaning users can deploy applications and process critical data on-site without latency issues from data transfers.

According to Gryf’s data sheet, the machine can provide up to 30 teraflops of FP64 performance from both Nvidia H100 NVL and H200 NVL, with 3.9Tbps and 4.8Tbps of bandwidth, respectively.

In addition to its “high-performance GPUs,” compute, storage, and network sleds, Gryf can be customized for specific workload demands. Users can also increase the performance of the system by stacking up to five units, interconnecting across the AI fabric, allowing any server to access any resource within the fabric as if it were on a single node.

“Gryf represents a fundamental shift in how organizations access and utilize high-performance computing at the edge,” said Alan Benjamin, CEO of GigaIO. “By bringing supercomputing capabilities to field operations in a portable form factor, we’re enabling real-time intelligence and analytics that were previously impossible without massive infrastructure. The strong interest from defense, intelligence, sports, media organizations, and the energy sector confirms the market need for this revolutionary approach to edge computing.”