What we know
An AMD executive highlighted that artificial intelligence will not replace high-performance computing (HPC). Despite AI’s rapid growth, high-precision workloads, especially those requiring double-precision floating-point (FP64) calculations, will remain essential for scientific and engineering applications. Researchers acknowledge that HPC’s role in handling complex, precise computations will persist alongside expanding AI use. Researchers agree AI will not supplant HPC. High-precision FP64 workloads will continue to be important even as AI adoption increases.
Why it matters
AI hardware often focuses on lower-precision calculations to maximize speed and efficiency, which limits its suitability for tasks requiring exact numerical accuracy. Many critical fields—such as climate science, physics, and computational chemistry—depend on the precision and rigor that traditional HPC provides. AMD’s perspective underscores the ongoing need for HPC systems, even as AI technologies advance and become more widespread.
What is still unknown
Details on independent evaluations of accuracy, the range of supported programming languages, and any planned release timelines were not provided. No independent tests of performance or precision have been reported.

