Keynote: Dr. James Chambers, Vice President of Silicon Engineering and Sourcing, NVIDIA
ISTFA Keynote
Dr. James Chambers
Vice President of Silicon Engineering and Sourcing
NVIDIA
The concepts of neural networks and artificial intelligence (AI) have been in development for decades, but it is only recently that NVIDIA and its partners have made them a reality at scale. By reinventing GPUs and computing systems, NVIDIA is leading the way in the Generative AI Era where every industry is racing to take advantage of this new capability. Today, NVIDIA has networked hundreds of CPU/GPU superchips together into a connected system acting as a single GPU computing platform for giant-scale AI and High Performance Computing applications. Taking advantage of the energy efficiency and throughput of accelerated computing, NVIDIA’s customers use these systems at data center scale to run their AI applications across multiple thousands of GPU chips running a single workload. This has led to more stringent quality and reliability requirements than ever before. In this talk, we will explore the implications the Generative AI Era is having on quality and reliability requirements, the types of latent defects that cause AI workloads to fail, future technology developments to support AI and their impact on failure analysis techniques and methodologies, and the application of AI to enhance the efficiency of failure analysis.