Jensen Huang desires to convey generative AI to each information middle Nvidia stated co-founder and CEO at this time at Computex in Taipei. In the course of the speech, Huang’s first public speech in almost 4 years, he stated, he made a slew of bulletins, together with the discharge of superchips and partnerships with main companies. Right here’s all of the information from the two-hour keynote.
- Nvidia’s GForce RTX 4080 Ti GPU for players is now in full manufacturing and being produced in “bulk portions” with companions in Taiwan.
2. Huang declares the Nvidia Avatar Cloud Engine (ACE) for video games, a customizable AI mannequin casting service with pre-trained fashions for recreation builders. It would give NPCs extra character by AI-powered language interactions.
3. Nvidia Cuda’s computing mannequin now serves 4 million builders and greater than 3,000 functions. Cuda noticed 40 million downloads, together with 25 million final yr alone.
4. Full quantity manufacturing of the HGX H100 GPU server has begun and is being produced by “firms throughout Taiwan,” Huang stated. He added that it’s the world’s first pc with a transformer motor.
5. Huang referred to Nvidia’s 2019 acquisition of supercomputer chipmaker Mellanox for $6.9 billion as “one of many largest strategic selections” it has ever made.
6. Manufacturing of the following era of Hopper GPUs will begin in August 2024, precisely two years after manufacturing of the primary era began.
7. Nvidia’s GH200 Grace Hopper is now in full manufacturing. The tremendous chip boosts 4 PetaFIOPS TE, 72 Arm CPUs linked by chip-to-chip hyperlink, 96GB HBM3 and 576 GPU reminiscence. Huang described because the world’s first accelerated pc processor that additionally has a large reminiscence: “that is a pc, not a chip.” It’s designed for extremely resilient information middle functions.
8. If the Grace Hopper’s reminiscence isn’t sufficient, Nvidia has the answer: the DGX GH200. It was created by first connecting eight Grace Hoppers along with three NVLINK switches after which connecting the pods along with 900 GB. Then lastly 32 are joined collectively, with one other layer of switches, to attach a complete of 256 Grace Hopper chips. The ensuing ExaFLOPS Transformer Engine has 144 TB of GPU reminiscence and features like a large GPU. Huang stated the Grace Hopper is quick sufficient to run the 5G stack in software program. Google Cloud, Meta and Microsoft would be the first firms to realize entry to the DGX GH200 and can examine its potentialities.
9. Nvidia and SoftBank have partnered to introduce the Grace Hopper superchip into SoftBank’s new distributed information facilities in Japan. They may be capable to host generative AI and wi-fi functions in a multi-tenant frequent server platform, lowering prices and vitality.
10. The SoftBank-Nvidia partnership will probably be based mostly on the Nvidia MGX reference structure, at present used at the side of firms in Taiwan. It offers system producers a modular reference structure to assist them construct over 100 server variations for AI, accelerated computing, and omniverse use. Firms within the partnership embrace ASRock Rack, Asus, Gigabyte, Pegatron, QCT and Supermicro.
11. Huang introduced the Spectrum-X accelerated community platform to spice up the velocity of Ethernet-based clouds. It contains the Spectrum 4 swap, which has 128 ports at 400 GB per second and 51.2 T per second. The swap is designed to allow a brand new sort of Ethernet, Huang stated, and is designed end-to-end to carry out adaptive routing, isolate efficiency and carry out in-fabric computing. It additionally contains the Bluefield 3 Good Nic, which connects to the Spectrum 4 swap to carry out congestion management.
12. WPP, the biggest promoting company on the planet, has partnered with Nvidia to develop a content material engine based mostly on Nvidia Omniverse. Will probably be in a position to produce pictures and video content material that can be utilized in ads.
13. Robotic platform Nvidia Isaac ARM is now accessible to anybody who desires to construct robots, and is full-stack, from chips to sensors. Isaac ARM begins with a chip known as Nova Orin and is the primary full reference stack for robotics, Huang stated.
Largely attributable to its significance in AI computing, Nvidia’s inventory has soared over the previous yr and at present has a market worth of round $960 billion, making it one of the vital worthwhile firms on the planet (solely Apple, Microsoft, Saudi Arabia). -Arabia). Aramco, Alphabet and Amazon are ranked larger).