Fifty Samsung presidents in one room outside Seoul, and the meeting ended without a supply contract.
Sam Altman sat with about 50 Samsung executives on Saturday at the company's training center in Yongin.
Samsung framed it as a leadership seminar.
His message was that adopting AI by itself gives nobody an edge, and that the payoff comes from moving faster and building better products.
The harder fact sits behind that meeting.
OpenAI's memory demand for Stargate has been projected at up to 900,000 DRAM wafers a month.
AMD's Lisa Su spent hours with Samsung's chip chief chasing HBM supply.
SK's chairman met Altman in Seoul the same day.
Micron's CEO expects the crunch to be worse in 2027 and 2028 than it is now.
Now picture the founder of a six person AI startup in Austin reading that news this morning.
She needs GPU memory to ship a model this quarter.
She is not on the invite list, and she is not going to be.
She does not need to be.
The compute she is waiting on already exists, on machines that can be rented for $0.15/hr instead of bought at shortage prices.
Buy Now Pay Later means zero upfront and payment after one month, which is the difference between testing the model this quarter and waiting until 2028.
What is the longest you have waited on a GPU or memory quote?
https://ai.bhkcloud.com/?utm_source=linkedin&utm_medium=social&utm_campaign=bhk_social_2026w45&utm_content=morning
Originally posted on linkedin_personal