THIS WEEK IN AI: ChatGPT Built its Own Chip, Micron Soars, a24's $75m Google Deal

June 26, 2026

AI Summary

5 min read

OpenAI's new AI chip, nicknamed Jalapeño, was designed in just nine months with help from the company's own models including ChatGPT and Codex.

OpenAI's move into custom silicon marks a shift toward full vertical integration. The chip is an ASIC built specifically for inference rather than a general-purpose GPU, addressing the memory bandwidth limits that currently leave Nvidia hardware underutilized during token generation. The design process itself relied on OpenAI's internal models and Codex to handle both software stack and hardware elements, shortening a timeline that normally requires one to two years. Partnerships with Broadcom, which already fabricates Google's TPUs, and MediaTek support the effort, and the first chips are expected by the end of the year. Hosts noted that this creates a potential flywheel: improved models accelerate chip design, while better hardware frees compute for further model training.

Continue reading the full summary in the app — free to try.

Read Full Summary →

Free • No credit card required

Listen to Audio Summary Open in App

Never miss an episode of Limitless Podcast

Get every new episode summarized in your inbox — free, ~5 minutes to read.

No spam. Unsubscribe anytime.

What you'll learn

1 (00:00) **OpenAI Announces Custom AI Chip** - First details on their in-house Jalapeño ASIC designed for OpenAI models
2 (00:36) **OpenAI Enters Hardware Stack** - Discussion of vertical integration from models to silicon
3 (01:25) **Memory Bandwidth as Core Bottleneck** - Why custom accelerators target underutilized GPU capacity
4 (02:21) **AI Models Designing Hardware** - Flywheel effect where better models accelerate chip design
5 (03:51) **Broader Compute Ownership Implications** - How the approach frees future compute for model training
6 (05:23) **Micron Earnings Surprise** - Massive beat with projected revenue exceeding Nvidia's prior-year Q3
7 (06:44) **AI Memory Demand Drivers** - Memory requirements for models, agents, and physical AI

+ Full timestamped outline available in the app

Show Notes

Today we're unpacking OpenAI’s plan to build its own custom AI chip—the Jalepeño—and what that means for its hardware strategy.

We also cover Micron’s strong earnings, its ties to Anthropic, Meta’s next model timeline, and new developments from Google, Amazon, Anthropic, and Valar Atomics.

------
🌌 LIMITLESS HQ ⬇️

NEWSLETTER: https://limitlessft.substack.com/
FOLLOW ON X: https://x.com/LimitlessFT
SPOTIFY: https://open.spotify.com/show/5oV29YUL8AzzwXkxEXlRMQ
APPLE: https://podcasts.apple.com/us/podcast/limitless-podcast/id1813210890
RSS FEED: https://limitlessft.substack.com/

------
TIMESTAMPS

0:00 OpenAI’s AI Chip Move
2:46 Compute and Hardware Strategy
5:30 Micron’s Earnings Surge
11:03 Meta’s Mythos Model Delay
14:47 A24 Meets Google AI
17:51 Sam Altman Movie Standoff
19:02 Claude Tag Enters Slack
21:10 Newsletter and Sponsorship Call

------
RESOURCES

Josh: https://x.com/JoshKale

Ejaaz: https://x.com/cryptopunk7213

------
Not financial or tax advice. See our investment disclosures here:
https://www.bankless.com/disclosures⁠