AEHR versus SMCI

The trends for both of these companies, like so many other semiconductor makers, are incredibly bullish. I am focusing on these two because they seem to me to have the most potential for further gains. I currently have about 10% in SMCI and 4% in AEHR. I’m worried about how far they have run, but given the demand for chips to power AI capabilities, I think they can go a lot higher. What I’m struggling with, though, is valuation, and which is the better investment, and I’m hoping to start a discussion about this here.

Here’s why I like SMCI:

  1. Valuation
    SMCI seems incredibly cheap to me if one considers that it trades at 21 times TTM earnings and 2 times sales over the same period and it is the FUTURE demand for its products that has ignited the rally in the stock. Twenty one times earnings is below the average multiple for a company in the S&P. But here we have a company selling one of the most in-demand products in the economy The company most associated with the surge in demand for chips to power AI, NVDIA, trades at 200 times TTM earnings, and 36 times sales. And NVIDIA is a nearly trillion dollar company by market cap compared to under $14B for SMCI.
  2. close association with NVIDIA. Again, NVIDIA is the poster child for the AI chip boom, and SMCI is a key supplier to NVDIA.

Here is why I like AEHR:

  1. It is also squarely in the middle of the AI-driven chip boom. Note the bullish commentary from the company cited by Saul that Every manufacturer of silicon carbide chips is currently at least in talks with us by now.
  2. market cap of just $1.2 billion.

My concern, though, is that 83 times TTM earnings and 19 times sales may be too pricey, especially for a hardware company.

I suspect my analysis is incomplete, and I’m hoping someone here will be so generous as to help me out :slightly_smiling_face: Thanks in advance!


A significant question regarding SMCI regards the sustainability of it’s growth.

Recent history has been great, but not so great if you look further back (without building a table as an example). All I can do at this point to a few things:

  • much more rapid fulfillment of sales in an environment demanding it compared to competition (anecdotally)
  • I’ve seen frequent comments about how good their product is and I’ve seen no comments to the contrary

The best proof of sustainability is the ongoing extension of that outsized performance… as in “We shall see”. Paying close attention to the news is prudent.

There is some criticism regarding margins. Sometimes they trim their margins even further to land key accounts. Regarding breast-beating over margins… I just point to the EPS, which, as you say, is pretty high compared to the share price… and it (the EPS) is growing quickly. $11/share in the fiscal year ending June 30.

Official management guidance tends to be conservative in some respects, but they ALSO make statements that shed light on their true thoughts. I’ll be paying close attention to the upcoming earnings statement (probably coming out in early August).

FWIW, I expect the upcoming year to have EPS (non-GAAP) in the $13-$15 range… which is a huge range compared to what appears to be $11 for this year. I’m guessing $14/share at the moment… and anyone can also guess at what an appropriate PE might be.

Given the not-so-solid foundation of guesses upon guesses ;)… I expect that SMCI should be valued considerably higher than the current share price and I’ll hold onto it in my “not a portfolio”.

He is no fool who gives what he cannot keep to gain what he cannot lose.


"Rosenblatt Securities analyst Hans Mosesmann on Tuesday initiated coverage of SMCI stock with a buy rating and a 12-month price target of 300. That would be a 34% increase from Monday’s closing price of 223.66.

AI now accounts for nearly 30% of Supermicro’s sales. Meanwhile, sales in that segment are accelerating with increasing visibility out one to two years, Mosesmann said.

Supermicro has competitive advantages thanks to innovations like liquid cooling at scale, he said. Further, the cooling technology prevents processors and servers from overheating.

“Liquid cooling at scale has been quite difficult to deploy given the complexity, expense, and reliability concerns (leaks or droplets),” Mosesmann said. “Supermicro’s liquid cooling technology, at scale, can increase rack compute power by over 2X (two times), which we see as a disruptive dynamic in a power-constrained data center.”"
Investors Business Daily

He is no fool who gives what he cannot keep to gain what he cannot lose.