In the USA, an AI supercomputer the size of a suitcase, Gryf, was created: what it can do
GigaIO, which develops complex infrastructure, has released a suitcase-sized artificial intelligence supercomputer called Gryf.
GigaIO, which develops complex infrastructure, has released a suitcase-sized artificial intelligence supercomputer called Gryf.
GigaIO, which develops complex infrastructure, has released a suitcase-sized artificial intelligence supercomputer called Gryf.
Gryf, jointly developed by GigaIO and SourceCode, is the «world’s first suitcase-sized AI supercomputer,» bringing data center-scale computing power directly to edge operations, enabling real-time insights and analytics.
GigaIO said the machine is designed and built in the US, and the company has already received «significant orders» from the US Department of Defense and the intelligence community. The Gryf can also support applications in healthcare, scientific research, industry and manufacturing, oil and gas, and sports analytics.
GigaIO says that Gryf is powered by FabreX AI technology, which delivers «unprecedented computing power in a rugged, field-ready design,» meaning users can deploy applications and process critical data on-site without data transfer delays.
According to the Gryf datasheet, the machine can provide up to 30 teraflops of FP64 performance thanks to Nvidia H100 NVL and H200 NVL graphics cards with bandwidths of 3.9 Tbit/s and 4.8 Tbit/s, respectively.
In addition to its «high-performance GPUs,» compute resources, storage, and network sleds, Gryf can be customized to specific workload requirements. Users can also increase system performance by stacking up to five units together, connecting them together via an 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 use 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 are enabling real-time intelligence and analytics that were previously impossible without massive infrastructure. The significant interest from defense, intelligence, sports, media, and energy organizations validates the market need for this revolutionary approach to edge computing.»


