Skip to content
Live
Trending storyDeveloping

AMD executive talks about AI and supercomputers

1 reports1 outlets2 hours agoUpdate

The story so far

AI overview

AMD executive Scott Tease said that AI will not replace traditional high-performance computing, and high-precision scientific computing will remain for a long time. AI is good at processing low-precision data, suitable for model training and rapid screening, but climate simulation, nuclear energy, and other cutting-edge research still need FP64-level numerical precision. AMD and IBM are advancing a pilot at Oak Ridge National Laboratory to explore the integration path of quantum error correction, AI-assisted scientific research, and traditional simulation. Even if AI applications continue to expand, high-precision scientific computing that relies on double-precision FP64 will remain for a long time. Future scientific research computing should not be 'AI replacing HPC,' but should build a heterogeneous computing system that collaboratively connects CPUs, GPUs, AI accelerators, traditional supercomputers, and quantum computers. Most HPC research institutions have realized that AI 'will never replace' traditional high-performance computing.

AI-written from the reports · Updated 2 hours ago

Coverage timeline

Follow the coverage to see every side of the story.

17 Nov
  1. ITHome · AI
    AMD executive: AI will not replace supercomputers, high-precision computing is still irreplaceable

    AMD executive Scott Tease says AI will not replace traditional high-performance computing, and high-precision scientific computing will remain for the long term. AI is good at processing low-precision data, suitable for model training and rapid screening, but climate simulation, nuclear energy, and other cutting-edge research still require FP64-level numerical precision. AMD and IBM are advancing a pilot at Oak Ridge National Laboratory to explore the integration path of quantum error correction, AI-assisted research, and traditional simulation.

Heat over time

Not enough data yet to draw a trend.