IBM is the first cloud customer for Intel’s Gaudi 3 AI accelerator chip, which it will make available in early 2025. The Gaudi 3 will be available for hybrid and on-site environments via the IBM Cloud, as part of Watsonx AI and on IBM data platforms. Gaudi 3, which began shipping in Q2 and is expected to go into mass production later this year, is IBM’s AI challenger to GPU accelerators from Nvidia and AMD, the latter having in January begun shipping its own HPC solution, the MI300X. Unlike that chip and Nvidia’s Hopper H100 and more recent Blackwell B200, the Gaudi 3 is not a GPU, but built on an architecture specifically for inference and deep learning. Continue reading IBM Cloud Is First to Widely Implement Intel Gaudi 3 AI Chips

Demand for artificial intelligence computer chips drove Nvidia income up 769 percent to nearly $12.3 billion for Q4, year-over-year, and 286 percent — to just over $29.7 billion — for the full-year fiscal 2024 frame that ended January 28. Revenue was $22.1 billion (+265 percent) and $60.9 billion (+126 percent) for the respective periods. Data center sales hit record highs of $18.4 billion for the quarter, up 409 percent from the previous year, $47.5 billion for the fiscal year, an increase of 217 percent. Gaming revenue was flat for Q4, at $2.9 billion, and up 115 percent for the year. Continue reading Nvidia Revenue and Profits Soar on Strength of AI Chip Sales

Nvidia is investing $30 billion in a new business unit focused on custom chips for high-performance computing. The company already controls about 80 percent of the advanced chip market but wants to avoid losing ground as alternatives spring up. Alphabet, AWS, Intel and AMD market high-end processors to third-parties, and Meta is expected to begin deploying its own Artemis AI chips this year. Nvidia has had discussions with Amazon, Google, Meta, Microsoft and OpenAI about helping them create bespoke chips and is also talking to automakers, cloud service providers (CSPs) and telecom companies, according to reports. Continue reading Nvidia to Launch Unit Devoted to Building Custom HPC Chips

Amazon Web Services, a leading provider of cloud computing services, is rolling out its new ARM-based Graviton3E chips for high-performance workloads, including tasks like weather forecasting and gene sequencing. AWS customers can rent the high-performance computing (HPC) power to take advantage of “performance gains and cost savings” as a result of making its own chips, Amazon says. The move makes AWS something of a competitor to other top chipmakers, including Intel, AMD and Nvidia, who continue to be among Amazon’s major chip suppliers. Amazon says it will continue to offer HPC services that rely on third-party chips. Continue reading AWS Touts HPC with Launch of Graviton3E Chip at re:Invent

Nvidia has begun previewing its latest H100 Tensor Core GPU, promising “an order-of-magnitude performance leap for large-scale AI and HPC” over previous iterations, according to the company. Nvidia founder and CEO Jensen Huang announced the Hopper earlier this year, and IT professionals’ website ServeTheHome recently had a chance to see a H100 SXM5 module demonstrated. Consuming up to 700W in an effort to deliver 60 FP64 Tensor teraflops, the module — which features 80 billion transistors and has 8448/16896 FP64/FP32 cores in addition to 538 Tensor cores — is described as “monstrous” in the best way. Continue reading Nvidia Touts New H100 GPU and Grace CPU Superchip for AI

Samsung has announced plans to build a $17 billion chip plant in Taylor, Texas. The news comes on the heels of a government push to jump-start more U.S. semiconductor manufacturing and Senate approval of $52 billion in industry subsidies for new processor factories. The South Korea-based electronics giant already operates a chip fabrication plant in Austin, Texas, opened in 1997 and expanded in 2007. The Taylor facility will create new sourcing for chips, which have become precious amidst a global shortage, although the new factory is not expected to become operational until 2024. Continue reading Samsung Plans to Construct a $17 Billion Chip Plant in Texas

Nvidia and the National Energy Research Scientific Computing Center (NERSC) debuted Perlmutter, an AI supercomputer that features 6,144 Nvidia A100 Tensor Core GPUs. Named for astrophysicist Saul Perlmutter, the supercomputer has been dubbed by Nvidia as “the fastest on the planet,” at processing with the 16-bit and 32-bit mixed-precision math used in AI applications, said the company’s global HPC and AI product marketing lead Dion Harris. Its first job will be to create the largest-ever 3D map of the visible universe. Continue reading Nvidia and NERSC Unveil a New Supercomputer for AI Tasks