First thread I’m starting here, so go easy on me ![]()
I’ll say up front this isn’t a classic Saul-style company. Gross margin is in the high 20s, there’s debt, it’s a mix of three businesses, and a lot of the growth comes from memory pricing. But the growth just turned sharply and I didn’t see a thread on it, so I’m curious what others think.
Penguin reported fiscal Q4 on Oct 6:
- Revenue $567M, up 68% YoY (Q3 was +48%, first half was roughly flat)
- FY26 revenue $1.73B, up 26%; non-GAAP EPS $2.87, up 51%
- FY27 guide about $2.43B, roughly +40%
The catch is where the growth comes from. Integrated Memory was $341M, up 158%, and is now about 60% of sales. Advanced Computing, the AI infrastructure business they want to be known for, was $154M and only up 11%, and down 14% for the full year because of the Penguin Edge wind down and lower hyperscale sales. The part they’re excited about, non-hyperscale AI infrastructure (enterprise, sovereign, neocloud), nearly doubled and is now two thirds of that segment, so there’s something real growing underneath.
The bull case as I see it: memory demand stays tight for a while, and by the time it cools, the AI infrastructure business is big enough to carry growth. They raised the FY27 Advanced Computing outlook from mid-teens to about 40% in one quarter, which helps that argument.
The bear case: it’s mostly a memory-cycle stock with low margins, and those businesses usually get valued that way once pricing turns. Two details from the release lean that way. The CFO said Q4 gross margin improved because memory got more profitable, partly offset by lower Advanced Computing margins, so the AI side isn’t carrying profits yet. And non-GAAP diluted shares went from 52.8M in Q3 to 62.0M in Q4, which I assume is tied to the convertible notes they issued in July. Worth keeping in mind before valuing it on EPS.
The thing I keep coming back to is how much of the memory growth is price versus volume. If it’s mostly price, the numbers could fade fast once the cycle turns, and then it comes down to whether the AI infrastructure business is big enough by then. Curious if anyone has a read on that.
No position. Mostly trying to decide if it belongs on my watchlist.