Nvidia’s ambitious future is to bring AI to every industry where GPU technology can be leveraged
why it matters: During the GTC 2023 keynote, Nvidia CEO Jensen Huang highlighted the breakthroughs of the new generation, which aims to bring AI to every industry. In partnership with tech giants such as Google, Microsoft and Oracle, Nvidia is making advances in AI training, deployment, semiconductors, software libraries, systems and cloud services. Other partnerships and developments announced include Adobe, AT&T, and automaker BYD.
Huang cited several examples of Nvidia’s ecosystem in action, including Microsoft 365 and Azure users gaining access to a platform for building virtual worlds, and simulations to train Amazon autonomous warehouse robots. using capabilities. He also noted the rapid rise of generative AI services such as ChatGPT, calling its success “AI’s iPhone moment”.
Based on Nvidia’s Hopper architecture, Huang announced a new H100 NVL GPU, which operates in a dual-GPU configuration with NVLink to meet the growing demand for AI and Large Language Models (LLM). The GPU features a Transformer engine designed for GPT-like processing models, which reduces LLM processing costs. Compared to the HGX A100 for GPT-3 processing, a server with four pairs of H100 NVLs can be up to 10 times faster, the company claims.
With cloud computing becoming a $1 trillion industry, Nvidia has developed the Arm-based Grace CPU for AI and cloud workloads. The company claims up to 2x performance over x86 processors with similar power in major data center applications. Then, the Grace Hopper superchip combines the Grace CPU and Hopper GPU to process the massive datasets typically found in AI databases and large language models.
Furthermore, Nvidia’s CEO claims that their DGX H100 platform, which houses eight Nvidia H100 GPUs, has become the blueprint for building AI infrastructure. Several major cloud providers, including Oracle Cloud, AWS and Microsoft Azure, have announced plans to adopt the H100 GPU in their offerings. Server makers such as Dell, Cisco and Lenovo are also building systems powered by the Nvidia H100 GPU.
Because clearly, generative AI models are all the rage, Nvidia is offering new hardware products, along with specific use cases to make inference platforms run more efficiently. The new L4 Tensor Core GPU is a universal accelerator optimized for video, delivering up to 120x better AI-powered video performance and 99% better energy efficiency than CPUs, while L40 graphics and AI-enabled graphics for image generation Optimized for 2D. , video, and 3D image creation.
Read also: Has Nvidia Conquered the AI Training Market?
Nvidia’s Omniverse is also present in the modernization of the auto industry. By 2030, the industry will mark a shift towards electric vehicles, new factories and battery megafactories. Nvidia says Omniverse is being adopted by major auto brands for a variety of tasks: Lotus uses it for virtual welding station assembly, Mercedes-Benz for assembly line planning and optimization, and Lucid Motors for creating digital stores with precise design data. does for. BMW partners with Idealworks for factory robot training and plans for an electric-vehicle factory entirely in the Omniverse.
Overall, there were too many announcements and partnerships to mention, but arguably the last big milestone came from the manufacturing side. Nvidia announced a breakthrough in chip production speed and energy efficiency with the introduction of “QLitho,” a software library designed to accelerate computational lithography by up to 40 times.
Jensen explained that the Qulitho could significantly reduce the extensive computation and data processing required in chip design and manufacturing. This will result in a significant reduction in power and resource consumption. TSMC and semiconductor equipment supplier ASML plan to incorporate CuLitho into their production processes.