Wang Long Chai
July 30, 2024 09:37
NVIDIA CEO Jensen Huang discusses advances in generative AI and accelerated computing at SIGGRAPH 2024, predicting a future where AI assistants are commonplace.
Speaking at the SIGGRAPH 2024 conference, NVIDIA CEO Jensen Huang shared his vision for a future where AI assistants become an integral part of everyday life and business operations. He highlighted the transformative potential of generative AI and accelerated computing technologies to revolutionize a wide range of industries.
Generative AI and Accelerated Computing
Huang highlighted the important advances in generative AI rooted in visual computing. He pointed out that accelerated computing not only increases productivity but also improves energy efficiency. According to Huang, “accelerated computing helps save 20 times, 50 times the energy even when performing the same processing.” He emphasized the importance of accelerating applications to reduce global energy consumption.
New announcements and integrations
The conference featured a number of announcements from NVIDIA. The company introduced its NIM family of microservices designed for a variety of workflows, including OpenUSD, 3D modeling, robotics, and industrial digital twins. These advancements are aimed at empowering developers, particularly by integrating Hugging Face Inference-as-a-Service on DGX Cloud.
Shutterstock also launched a Generative 3D service, and Getty Images upgraded its product with NVIDIA Edify technology. In AI and graphics, NVIDIA released new OpenUSD NIM microservices and reference workflows tailored for generative physical AI applications, including a program to accelerate the development of humanoid robots.
Practical applications across industries
Huang and WIRED’s Lauren Good discussed the evolution of visual computing from computer games to digital animation and GPU-accelerated computing, leading to the generative AI that powers today’s industrial-scale AI factories. Huang explained that robotics requires advanced AI and realistic virtual worlds for training before deploying them on next-generation humanoid robots.
He also noted that a range of industries, from scientific computing to industrial visualization, will be transformed by generative AI. For example, NVIDIA’s Omniverse system, built on the OpenUSD standard, will leverage generative AI to create assets for major brands.
The Future of AI Assistants and Digital Agents
Huang predicted that digital agents and AI assistants will augment every job within a company. He suggested a future where humans work alongside AI, citing customer service as one of the most promising use cases. “In the future, it will still be humans, but there will be AI in the loop,” Huang said.
All of these developments promise to amplify human productivity and creativity. Huang concludes by imagining the compelling stories that could be told using these transformative tools.
For more information, visit the NVIDIA blog.
Image source: Shutterstock