Three things happened to OpenAI last week. The company went from being a model lab to something closer to a vertically integrated empire that wants to design its own silicon, own a chunk of the world’s memory supply, and cure the common cold.
John Coogan, the co-host of TBPN walked through all of it, and the through-line is that Sam Altman’s organization is no longer satisfied renting infrastructure from anyone, including biology.
The Jalapeño chip and a compressing design cycle
The headline move is “Jalapeño,” a custom chip OpenAI designed with Broadcom, described as “purpose-built for LLM workloads powering ChatGPT, Codex API, future agentic products.” Coogan noted that many are calling it “the first chip designed with AI agents in the loop so you can write the instruction set much quicker.” The model is helping design the chip that runs the model.
Coogan also picked up on the timeline. “For a long time it was like, oh, it’ll take 5 years to launch a chip and now companies are clearly doing it faster,” he said, while the caller on the segment added that “you gotta give a lot of credit to the humans on both sides.”
Broadcom (NASDAQ:AVGO | AVGO Price Prediction) is the obvious beneficiary here. Hock Tan’s company already runs the custom ASIC playbook for Google’s TPU and last quarter reported AI semiconductor revenue of $10.80 billion, up 143% year-over-year, with Q3 guidance pointing to roughly $16.00 billion in AI semi revenue, more than 200% growth. CLSA recently called out “clarity in Broadcom’s custom ASIC roadmap” with “potential 30% upside to AI semiconductor revenue guidance.” You can see the receipts in Broadcom’s own Q2 8-K filing. Shares have had a rough month, down 13.33% over the trailing thirty days, though the one-year picture remains a 36.1% gain.
Locking up 40% of the world’s DRAM wafers
Then there is the memory story, which is arguably wilder. Per Coogan, OpenAI reportedly secured deals to purchase “40% of the global raw undiced DRAM wafer output until 2029.” He framed it as a real supply-chain play, where you grab raw materials that “cannot be used until they’re processed” and sit on them through the decade.
That move lands directly on Micron Technology (NASDAQ:MU), whose Q3 FY2026 revenue more than quadrupled year-over-year and whose CEO Sanjay Mehrotra said “Micron’s record fiscal Q3 financial results and even stronger outlook for Q4 reflect the strategic value of memory in the AI era.” Micron’s latest print is staggering on its own terms. Revenue of $41.46 billion, gross margins of 84.6%, Q4 guidance calling for roughly $50 billion in revenue. Wall Street targets from TD Cowen, Cantor Fitzgerald, and Susquehanna run up to $1,750, with analysts arguing AI has created a “permanent, higher demand baseline, fundamentally changing the historically cyclical memory business.”
Retail noticed. On Reddit, the dominant narrative this week was “MU $2000 is no longer a meme,” with sentiment scores swinging from a pre-earnings 38 (bearish) to a post-earnings 81 (very bullish). The stock is up roughly 800.86% over the past year.
A $500 million moonshot to end the common cold
OpenAI, Anthropic, Stripe, and Bill Gates are jointly putting $500 million into a new organization called Intercept, whose goal is to “prevent the common cold and the flu and eventually eliminate all respiratory viruses completely.” Coogan added the relatable detail, citing a chart showing parents with multiple kids are sick up to 56% of the year.
The backer list is the interesting part. Two foundation-model rivals, a payments company, and the world’s most famous public-health philanthropist all writing checks into the same vehicle. That suggests Altman views AI-driven drug discovery as either a genuine moonshot or a useful brand halo, possibly both.
Stack the three threads together and you get a company building its own silicon, hoarding the raw memory wafers needed to run that silicon, and bankrolling a biology shop on the side. Whoever supplies OpenAI now supplies a customer that wants to control the entire stack, including, apparently, your immune system.