DOOMBENCH.COM
Competitive race

xAI brings its 100,000-GPU Colossus cluster online

xAI brought Colossus online with 100,000 NVIDIA H100 GPUs to train Grok models after a roughly four-month build, creating the largest known single AI training cluster at the time.

0 comments · 0 votesOpen discussion

Public discussion is readable by everyone. Sign in to comment, reply, or vote.

No comments yet. Start the discussion.

Send detailed or private feedback →
CURRENT ASSESSMENT · REVISION 1
TOWARD DOOM68confidence 94/100

Why it moved the index

A 100,000-GPU training system materially expanded one frontier laboratory's scaling capacity and intensified compute competition, with later NVIDIA evidence independently confirming the cluster's scale and use for Grok training.

Assessment history

  1. R1
    Toward 68 · confidence 94

    New September 2024 frontier-compute milestone with later first-party hardware confirmation and no durable collision.

    12 Aug 2026

Share this page

DoomBench social sharing card for xAI brings its 100,000-GPU Colossus cluster online.

  1. DoomBench assesses “xAI brings its 100,000-GPU Colossus cluster online” as evidence moving toward doom, with magnitude 68 and confidence 94 out of 100 in the competitive race category.

  2. The DoomBench assessment of “xAI brings its 100,000-GPU Colossus cluster online” is based on reporting from Fortune and records the editorial rationale, source quality, attribution, and revision history.

  3. DoomBench summarizes “xAI brings its 100,000-GPU Colossus cluster online” as follows: xAI brought Colossus online with 100,000 NVIDIA H100 GPUs to train Grok models after a roughly four-month build, creating the largest known single AI...

    https://www.doombench.com/news/xai-brings-its-100-000-gpu-colossus-cluster-online-2024-09-03