GTC 2023: AI's iPhone moment

Nvidia held the GTC conference on the evening of March 21, 2023. At the conference, Mr. Huang Renxun shared the latest developments and future trends of Nvidia's latest products and the AI field. Over the past month, major manufacturers have released and upgraded large-scale models and related applications intensively. At present, the AIGC industry has gone from a single phenomenon-level product to a major turning point that is expected to drive the productivity revolution in various industries.

Nvidia has launched a new GPU inference platform (L4, L40, H100 NVL, Grace Hopper chip) with four configurations for accelerating AI video, image generation, large-scale language model deployment, and recommendation systems. Among them, the DGXH100, which is suitable for the training phase, has 8 H100 GPU modules, which can provide 32PetaFLOPS computing power at FP8 precision, and provide a complete NVIDIA AI software stack to help simplify AI development. The NVIDIA DGXH100 AI supercomputer is currently in full production. Many companies in the field of generative AI, including cloud computing giants, are using the H100 GPU provided by Nvidia to accelerate their work. Accelerated computing products and the development of the AI industry coexist and co-prosper. Nvidia continues to provide a more powerful computing power base for larger-scale AI model training, which has played an important role in promoting the cutting-edge development of AI training and reasoning. The booming AI boom has brought a broader market and opportunity.

Nvidia also announced the joint launch of DGXCloud service with Google Cloud, Microsoft Azure, and Oracle Cloud. DGXCloud gives customers instant access to NVIDIA AI supercomputing in the cloud at a global scale. Enterprises can rent DGXCloud clusters on a monthly basis to quickly and easily scale the development of large multi-node training workloads without waiting for accelerated computing resources that are often in high demand. DGXCloud, which pays $37,000 a month, allows companies to train their own ChatGPT on the web, allowing commercial companies to use super computing power and AI application libraries faster and more conveniently.

Nvidia announced a new technology called CuLitho lithography at the GTC conference to speed up the design and manufacture of semiconductors. The software uses Nvidia chips to speed up the steps between software-based chip design and the physical fabrication of the photolithography masks used to print that design on the chip. The advance would allow chips with smaller transistors and circuits than today, while speeding up time to market and improving the energy efficiency of the massive data centers that run around the clock to power the manufacturing process. In addition, NVIDIA has also released a series of new developments in the fields of automobiles, quantum computing, and medical biology, including the opening of PaaS service NVIDIA Omniverse Cloud to specific enterprises, expanding cooperation with BMW Group to build a virtual factory, and cooperation with BYD on the NVIDIA DRIVE Orin platform.