Today, NVIDIA announced an expedited timeline for product releases along with the widespread use of the NVIDIA SpectrumTM-X Ethernet networking platform.
Among the first AI cloud service providers to adopt NVIDIA Spectrum-X to deliver exceptional networking performance to their AI infrastructures are CoreWeave, GMO Internet Group, Lambda, Scaleway, STPX Global, and Yotta. Furthermore, a number of NVIDIA partners have revealed Spectrum-based products; these include ASUS, GIGABYTE, Ingrasys, Inventec, Pegatron, QCT, Wistron, and Wiwynn. These companies join Dell Technologies, Lenovo, Hewlett Packard Enterprise, and Supermicro in integrating the platform into their products.
“Every company must prioritize networking innovation to gain a competitive edge,” stated Gilad Shainer, senior vice president of networking at NVIDIA, “as fast advancements in revolutionary technologies like generative AI underscore this point.” In order to enable companies to fully utilize the potential of their AI infrastructures to alter their industries and processes, NVIDIA Spectrum-X is revolutionizing Ethernet networking.
The Annual Release Cadence
With 1.6x faster generative AI network performance than conventional Ethernet fabrics, Spectrum-X is the world’s first Ethernet fabric designed with AI in mind.
NVIDIA founder and CEO Jensen Huang announced today during his COMPUTEX keynote in Taiwan that NVIDIA plans to launch new Spectrum-X products every year, delivering increased bandwidth and ports as well as enhanced software feature sets and programmability to drive industry-leading AI Ethernet networking performance. This is in response to the industry’s strong demand for the performance that Spectrum.
Spectrum-X, an end-to-end platform designed with the performance and functionality needed by generative AI clouds, including the NVIDIA Spectrum SN5600 Ethernet switch and the NVIDIA BlueField® -3 SuperNIC. It provides the highest-performance Ethernet networking for AI, delivering predictable results for thousands of concurrent AI operations at any scale by utilizing adaptive routing and congestion control for maximum bandwidth and noise isolation.
Combined with NVIDIA BlueField-3 DPUs, the platform enables advanced cloud multi-tenancy, GPU compute elasticity and zero-trust security. With it, cloud service providers can accelerate the development and deployment of AI solutions while improving their return on investment.
During Huang’s COMPUTEX keynote speech, discover more about NVIDIA Spectrum-X.