---
format: "aidr-story-markdown/v1"
id: "3ef6cc6787d08ce73da363e669a66bc735ccb77576eedbfaaedcce2da6bab1eb"
canonical_url: "https://aidr.today/3ef6cc67?lang=en"
title: "OpenAI AI-Designed Jalapeño Chip Reaches Tapeout in 9 Months With AI Code 1.8x Faster"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-08-31T18:48:23.000Z"
category: "Chips"
topics: ["openai","chips","inference","agent","apple"]
source_urls: ["https://huggingnews.com/ai/update-openai-ai-designed-jalapeno-chip-reaches-tapeout-in-9-months-with-5ce6e29e","https://x.com/LearnOpenCV/status/2094419249636716593","https://x.com/openlabxorg/status/2094313418459402435"]
summary: "AI-authored code for specific attention and MoE blocks ran 1.5 to 1.8 times faster than expert human code on the new Jalapeño chip. OpenAI used AI to move the hardware from design to tapeout in 9 months, building the architecture to be easily programmable by AI rather than human experts. The company also purchased thousands of Mac minis and Mac Studios to train AI agents using reinforcement learning. Apple's unified memory architecture supports these workloads, which require agents to interact with real operating systems and GUIs, differing from traditional model pretraining."
---

# OpenAI AI\-Designed Jalapeño Chip Reaches Tapeout in 9 Months With AI Code 1\.8x Faster

> [Open the canonical story](<https://aidr.today/3ef6cc67?lang=en>)

**Published:** 2026-08-31T18:48:23.000Z
**Category:** Chips
**Topics:** openai, chips, inference, agent, apple

## Summary

AI\-authored code for specific attention and MoE blocks ran 1\.5 to 1\.8 times faster than expert human code on the new Jalapeño chip\. OpenAI used AI to move the hardware from design to tapeout in 9 months, building the architecture to be easily programmable by AI rather than human experts\. The company also purchased thousands of Mac minis and Mac Studios to train AI agents using reinforcement learning\. Apple's unified memory architecture supports these workloads, which require agents to interact with real operating systems and GUIs, differing from traditional model pretraining\.

## Sources

- [Story source](<https://huggingnews.com/ai/update-openai-ai-designed-jalapeno-chip-reaches-tapeout-in-9-months-with-5ce6e29e>)
- [Supporting source](<https://x.com/LearnOpenCV/status/2094419249636716593>)
- [Story source](<https://x.com/openlabxorg/status/2094313418459402435>)