Now Playing
Samsung vs TSMC: OpenAI's Foundry Shift, HBM4E & the Packaging Crunch
Audio is available on Spreaker — see link below.
OpenAI's Gen Two chip is in evaluation at Samsung's SF2P node — a direct challenge to TSMC's grip on AI logic silicon. Today's briefing covers TSMC's record $16.35B August revenue, Samsung HBM4E samples, Positron's $875M LPDDR5X bet, and Marvell's XPU ramp.
Audio is available on Spreaker — see link below.
OpenAI is moving to break its dependence on TSMC for AI logic silicon, and the vehicle is a foundry partnership with Samsung built on Broadcom's two-hundred-billion-dollar manufacturing MOU. This is the clearest signal yet that the AI chip supply chain is entering a multi-foundry era.
Meanwhile, TSMC just posted August revenue of sixteen-point-three-five billion dollars, up fifty-three percent year over year. That's the fourth consecutive monthly record.
On memory, Samsung has shipped its first HBM4E samples. Projections put Samsung's HBM market share at roughly forty percent by the fourth quarter of twenty-twenty-six, up from thirty-three percent in Q2.
Not everyone is chasing HBM. Positron raised eight-hundred-and-seventy-five million dollars in a Series C led by NEA, reaching a five-billion-dollar valuation on a chip that deliberately avoids HBM entirely.
Two other developments sharpen the picture. Marvell expects its XPU business to more than double in fiscal twenty-twenty-eight, with acceleration into twenty-twenty-nine.
The signal to track from here is straightforward. Watch whether OpenAI's Samsung evaluation converts to a signed contract, and watch whether SF2P yield holds at scale when production volumes arrive.
Chapter summary auto-generated from the verified script. Listen to the full episode for the complete content.