Samsung Just Printed $58B in One Quarter. Here Is What It Means for Every Memory Name You Hold

Samsung released preliminary Q2 guidance today: operating profit around 89 trillion Korean won, which converts to roughly 58 billion dollars. That is up about 1,810 percent year over year, the highest quarterly operating profit ever recorded by any tech company, topping Nvidia's best quarter and Apple's best quarter. The market is talking about whether the AI memory supercycle is over. Samsung just answered the question.

What a $58 billion quarter actually proves

You do not print the largest quarterly profit in tech history on weak demand. Samsung's number was driven by HBM pricing and DRAM pricing staying at elevated levels even as the glut narrative ran wild in the headlines. When one of the three major memory makers posts numbers this large, it does not happen because AI datacenters suddenly stopped ordering. It happens because they could not get enough and the pricing held.

The short thesis got the number wrong

The bear case I have heard all month is that supply is catching up, NAND flash spot prices will roll over, and memory follows the old commodity cycle into a bust. NAND flash spot prices are up about 670 percent into 2026. Micron guided around 50 billion dollars for Q4 revenue. Now Samsung posts 58 billion for a single quarter. Three data points all pointing the same direction is not a coincidence.

Where I am positioned and why

I own $Micron Technology (MU.US)$ as my core US memory exposure, cost basis around 1,050 after adding on the break below a thousand last week. I added a little back because of exactly this kind of catalyst. NAND saw a breakthrough last week on the manufacturing side as well, which is the longer term cost story. And on Friday, SK Hynix is doing its Nasdaq debut, targeting around 29 billion dollars raised, the largest foreign listing in US history. Three memory companies all spending billions to expand capacity at the same time is the loudest possible signal about where the demand goes.

The one risk I actually respect

NAND flash went up 670 percent. At some point that attracts overcapacity and the cycle turns. I am not a perma-bull on memory. But the tell for me is hyperscaler capex guidance. When Nvidia or the cloud trio signals a meaningful pause, I trim and ask questions later. Until that data point arrives, I hold the winners and use the volatile days as buy signals, not exit signals.

Not financial advice, just how I read the setup.