OpenAI Bypasses Nvidia With Ultra-Fast Codex on Cerebras Wafer-Scale Chips
TL;DR Summary
OpenAI released GPT-5.3-Codex-Spark, a fast, text-only coding model that runs on Cerebras Wafer-Scale Engine 3 and achieves about 1,000 tokens per second, roughly 15x faster than its predecessor and faster than Nvidia-based options; it's a research preview for ChatGPT Pro and select partners, with a 128k-token window, built for speed-over-depth coding tasks, signaling OpenAI’s push to diversify hardware away from Nvidia.
- OpenAI sidesteps Nvidia with unusually fast coding model on plate-sized chips Ars Technica
- OpenAI Debuts First Model Using Chips From Nvidia Rival Cerebras Bloomberg.com
- GPT-5.3-Codex is now generally available for GitHub Copilot The GitHub Blog
- Opus 4.6, Codex 5.3, and the post-benchmark era Interconnects AI
- Introducing OpenAI GPT-5.3-Codex-Spark Powered by Cerebras Cerebras
Reading Insights
Total Reads
0
Unique Readers
7
Time Saved
5 min
vs 6 min read
Condensed
94%
1,051 → 61 words
Want the full story? Read the original article
Read on Ars Technica