Nvidia enabling AI factories

ASRock Rack, ASUS, GIGABYTE, Ingrasys, Inventec, Pegatron, QCT, Supermicro, Wistron and Wiwynn will deliver cloud, on-premises, embedded and edge AI systems using Nvidia GPUs and networking, says Nvidia.

“The next industrial revolution has begun,” says  Huang, “companies and countries are partnering with Nvidia to shift the trillion-dollar traditional data centres to accelerated computing and build a new type of data centre — AI factories — to produce a new commodity: artificial intelligence.”

The offerings will range from single to multi-GPUs, x86- to Grace-based processors, and air- to liquid-cooling technology.

Additionally, to speed up the development of systems of different sizes and configurations, the Nvidia MGXTM modular reference design platform now supports Nvidia Blackwell products.

This includes the Nvidia GB200 NVL2 platform, built to deliver unparalleled performance for mainstream large language model inference, retrieval-augmented generation and data processing.

GB200 NVL2 is suitable for emerging market opportunities such as data analytics, on which companies spend tens of billions of dollars annually.

Taking advantage of high-bandwidth memory performance provided by NVLink-C2C interconnects and dedicated decompression engines in the Blackwell architecture, speeds up data processing by up to 18x, with 8x better energy eciency compared to using x86 CPUs.

To meet the diverse accelerated computing needs of the data mcentres, Nvidia MGX provides computer manufacturers with a reference architecture to quickly and cost-eectively build more than 100 system design congurations.

Manufacturers start with a basic system architecture for their server chassis, and then select their GPU, DPU and CPU to address different workloads. To date, more than 90 systems from over 25 partners have been released or are in development that leverage the MGX reference architecture, up from 14 systems from six partners last year.

AMD and Intel are supporting the MGX architecture with plans to deliver, for the first time, their own CPU host processor module designs. This includes the next-generation AMD Turin platform and the Intel® Xeon 6 processor with P-cores (formerly codenamed Granite Rapids).

Any server system builder can use these reference designs to save development time while ensuring consistency in design and performance.

Nvidia’s latest platform, the GB200 NVL2, also leverages MGX and Blackwell. Its scale-out, single-node design enables a wide variety of system congurations and networking options to seamlessly integrate accelerated computing into existing data center infrastructure.

The GB200 NVL2 joins the Blackwell product lineup that includes Nvidia Blackwell Tensor Core GPUs, GB200 Grace Blackwell Superchips and the GB200 NVL72.

Nvidia’s  partner ecosystem includes TSMC, the world’s leading semiconductor manufacturer and an NVIDIA foundry partner, as well as global electronics makers, which provide key components to create AI factories.

These include manufacturing innovations such as server racks, power delivery, cooling solutions and more from companies such as Amphenol, Asia Vital Components (AVC), Cooler Master, Colder Products Company (CPC), Danfoss, Delta Electronics and LITEON.

As a result, new datacentre infrastructure can quickly be developed and deployed to meet the needs of the world’s enterprises — and further accelerated by Blackwell technology, Nvidia Quantum-2 or Quantum-X800 InniBand networking, Nvidia SpectrumTM-X Ethernet networking and Nvidia BlueField-3 DPUs — in servers from leading systems makers Dell Technologies, Hewlett Packard Enterprise and Lenovo.

Enterprises can also access the Nvidia AI Enterprise software platform, which includes Nvidia NIMTM inference microservices, to create and run production-grade generative AI applications.