With the stock of many leading technology companies rocketing to all-time highs recently, are there any values left in the sector?
One part of the tech sector in a big downturn right now is chip companies that enable electrification. That's because the growth of electric vehicles (EVs) slowed down in a big way last year.
However, not only is the following supplier of semiconductor testing equipment a way to play the EV rebound, but its equipment also could eventually find its way into even more tech applications in the back half of the decade -- including AI.
Aehr Test Systems helps limit defects in hard-to-make chips
Aehr Test Systems (AEHR 0.54%) has seen its stock swoon 78% from all-time highs, with its market cap falling to just $332 million as of this writing.
Aehr's revenue boomed over 2022 and early 2023, as its wafer-level test and burn-in equipment was adopted to test silicon carbide chips. Silicon carbide (SiC) is a novel alloy material that is just now being used to manufacture power semiconductors.
Power semiconductors have a range of applications, but their biggest growth market is powering drivetrains of EVs.
SiC has properties that make it favorable for EVs -- it has better conductivity and heat dissipation properties than traditional silicon. That's especially useful in chips that need to run at high temperatures, such as in automobiles or other mission-critical industrial machines. SiC chips therefore increase the performance and range of battery EVs.
The problem is that silicon carbide is somewhat difficult to produce without defects. But Aehr's testing equipment allows chip companies to test semiconductors when they are still on the wafer, subjecting the chips to elevated voltages and/or temperatures for 2 to 48 hours, a process called "burn in." This allows chipmakers to test a chip to see if it fails early in its life, which is when the vast majority of faulty chips fail.
But new applications mean risks at the start
So what has gone wrong for Aehr? For one, the SiC industry is still in its infancy, with industrial chipmakers only just now ramping up large plants and figuring out how to produce SiC at scale. And because Aehr's equipment is used in testing after production, it only goes into a fab once all of the other equipment has gone in and begins producing chips.
Thus, Aehr's revenue is highly concentrated in just a few customers and is subject to lumpy orders. The company claims it has seven silicon carbide customers today, but only a few -- those that have managed to begin producing SiC chips in material quantities -- are currently ordering equipment.
In fact, in the year-ago quarter, one customer accounted for a whopping 82.5% of the company's sales. As that customer slowed orders amid the current EV slowdown, Aehr saw a 66% decline in revenue last quarter. Ouch.
Image source: Getty Images.
But a recovery could be massive
Given that just about every EV maker is slowing down production amid the industrywide decline, it's no surprise to see a violent correction for Aehr's sales. But more SiC power-chip adoption will extend the range for EVs, helping customers potentially get over their "range anxiety" that could be limiting the current audience for pure battery-powered EVs.
Of course, there is also a school of thought that alternatives to EVs might emerge, putting the entire EV market at risk. But even potential alternatives are likely to come in a hybrid form, which would still have a battery and thus also use some silicon carbide chips. While the SiC content won't be quite as high in hybrids, they would still see a big SiC content increase in the future.
Test and burn-in for other applications?
But the biggest upside scenario for Aehr is if wafer-level test and burn-in is adopted in other tech applications outside of EV-focused SiC chips. To that end, the company already has engagements with customers that make memory chips -- gallium nitride (GaN) chips -- which are used in things like phone chargers and solar inverters, silicon photonics (where an optical transceiver is packaged together with a silicon chip, for uses like optical data transmission), and even AI accelerators.
Prior to today, many device makers would test chips at the device level, rather than the wafer level, and then just eat the small loss in yield that happens when they had to throw away an entire device. However, with chips now made with ever smaller and smaller transistors on larger die sizes, and many chips now made out of "chiplets" produced in entirely different plants before being stitched together, there are more opportunities for defects. That could spur wafer-level test and burn-in for more applications.
In a recent investor presentation, Aehr noted several new engagements in new industry verticals, including a large flash-memory manufacturer. This is on top of existing engagements with silicon photonics manufacturers and GaN clients.
But the most enticing newly announced engagement is with an unnamed "AI accelerator" company. This company is currently testing Aehr's Fox-XP product for wafer-level burn-in.
Assuming the test is successful, Aehr says this company could adopt its equipment more broadly. Given the hypergrowth that AI accelerators are seeing today -- a market that currently produces no revenue for Aehr -- this would be the biggest new opportunity.
High uncertainty, but massive upside
No doubt, Aehr's financials look ugly right now. But even in the recent quarter, which saw a dramatic decline in revenue, it had only a $1.5 million operating loss, and the company still made a $9.3 million operating profit over the past three quarters. Meanwhile, it is well capitalized for its small size, with $47.6 million in cash and no debt.
That should give the company plenty of time to weather this downturn and make it to the next up cycle. Given the huge decline in the stock and the intriguing potential for wafer-level testing, aggressive investors might want to buy the unloved Aehr. Years from now, you could be very glad you did.
