Skip to content
NewA calmer, faster ZeroHour

Anthropic says Claude Mythos cut AMD GPU porting to three days

Original title: Claude Mythos slashed software work to onboard AMD GPUs to around three days

Hacker News · AIby nsoonhui ()1 min read2
Part of a story covered by 3 sourcesAnthropic says Mythos cut AMD GPU porting to about three daysMerged summary & timeline
AI summary · x-ai/grok-4.7Summaries by Vani AI

The claim covers prompt processing on MI355X GPUs, with Mythos writing almost all kernels and little human help.

Anthropic told IPO investors that Claude Mythos Preview did the software work to run one of its models on AMD MI355X GPUs in about three days, a job it says previously took six to 12 months for dozens of engineers. The company said the work reached a useful performance level with basically no human intervention and that Mythos wrote almost 100% of the kernels. Anthropic already runs models on Nvidia and AMD GPUs, Google TPUs and AWS Trainium.

Unconfirmed: The timeline and kernel-authorship figures are Anthropic's statements to IPO investors, not an independent benchmark.

Key facts
  • Tom Brown, Anthropic co-founder, made the time comparison to investors.
  • The three days cover prompt processing at a useful performance level.
  • Anthropic says Mythos ported or wrote almost 100% of the kernels.
  • Anthropic manages its multi-chip fleet as one fungible compute pool.
Explain this AI news with Vani AIWhat was announced, what is new and why it matters, in a new chat

From the original article

1 min read at tomshardware.com

Anthropic told IPO investors that its Claude Mythos Preview model did the software work required to get its AI running on AMD’s GPUs in about three days, with “basically no human intervention,” a job it says used to take six to 12 months. The three days cover prompt processing for one Anthropic model on AMD MI355X GPUs “at a useful level of performance,” with Mythos porting or writing “almost 100% of the kernels.”

Hacker News · AI

By nsoonhui · tomshardware.com

Read the full article at Hacker News · AI

A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.