After reviewing the latest earnings news and the guidance, I think it's probably a good idea to buy SIMO here and hold it for at least two years.
Q4 2025 sales were a company record, and sales are going up. It's just starting.
90% of SIMO's current business is controllers for gadgets/tablets/smart "this and that" (eMMC), cell phones (UFS) and PCs (client SSDs).
SIMO is a monopoly in eMMC controller chips. 800 million eMMC controllers are sold each year, and the number is growing slowly as new devices need basic NAND storage. Samsung was the market share leader, but they have stopped development and are generally exiting the market. SIMO is the only meaningful merchant maker of eMMC controllers.
In cell phones (UFS) SIMO used to compete with the NAND makers internal controller groups for design slots. Two NAND makers (Micron and perhaps Samsung (!!)....) have de-emphasized UFS controllers for cell phones. Micron has exited UFS controllers completely. Samsung has de-emphasized the low end of UFS. When they're not selling UFS controllers what they do is make a NAND memory wafer (lots of chips) and sell it to another entity (electronics module maker, cell phone maker) and that entity has to buy a controller from someone and package it together into a module for use in a cell phone. Those customers (module makers or cell phone makers) don't make NAND controllers, they need to buy one, most likely from SIMO. The size of the market (in units) that Micron and Samsung are leaving is huge compared to SIMO's current market share in cell phones. Further, since cell phone controllers are now LOW GROSS MARGIN compared to NAND memory, other NAND makers may do the same thing (exit the low end cell phone controller business). The NAND makers can get a 60%+ gross margin on the memory wafer, and probably a 25% gross margin on the controller portion of the module. And if they sell the module to the cell phone maker they need customer support, design assistance, etc etc, all of which is costs. If they just make and sell the memory wafer all those costs go away, and their profit margin (in this segment) goes up quite a lot. With NAND memory gross margins so high it completely makes sense that ALL the NAND makers want someone else to do the low margin business of making controllers and doing all the design / sales / integration of the memory + controller module. SIMO will do it. That is exactly SIMO's business. This segment is going to grow, like crazy, for a few years, like at least 3, and if more NAND makers exit this business, then as many years as it takes for SIMO to get all of it.
In PCs (client SSD) the business is similar to cell phones, but perhaps 2-3 years behind cell phones in this transition from internal NAND controller development by the NAND makers toward mostly outsourced. There is currently a technology transition going on in client SSD. The main connection protocol between controller and memory is called PCIe gen4. PCs are beginning to transition to PCIe gen5 and that transition will occur over the next 2-2.5 years, first in the high end PC, and moving down the tech ladder as time progresses. SIMO thinks they currently have 30% global market share in client SSD, but they think they have more than 50% of design wins in PCIe gen5. As the tech transition from gen4 to gen5 occurs in the next few years, SIMO's sales will grow with the larger design win market share. And then in about 2030 the PC industry will move from gen5 to gen6. In that gen5 to gen6 transition it is easy to see the NAND makers doing the same thing then that they are doing now in cell phone controllers - exit the segment. It's a low end mass market consumer segment, with price sensitivity and low gross margins relative to the rest of their business, and there is an adequate alternative maker of gen6 PC controllers - SIMO. This indicates this segment - PC client controllers, about 60% of SIMO's current business - will grow sales for perhaps 5 years.
That's 90% of the SIMO story. It's their LEGACY business. It's going to grow regardless of trends in the cell phone and PC industry because the growth is not due to industry growth, it's due to market share gains from NAND makers that don't want or need to be in the low end to make a shit ton of money.
What's different from SIMO's past (where Lucy always pulled the football at the last moment)? Previously (from 2010 to 2024) SIMO' cell phone controller business sold through NAND makers in an OEM relationship (mostly in cell phone controllers), and took the NAND makers unwanted design slots in client SSD PC controllers. Every few years a NAND maker OEM relationship (which had become very large) collapsed. In the tech transition from eMMC to UFS in cell phones, SIMO lost SK Hynix as a partner as SK used SIMO for eMMC controllers but went internal with UFS controllers (so all of SIMO's sales in this segment went to zero). In the transition from UFS 3.0 to UFS 4.0 Micron stopped using SIMO UFS 3.0 controllers in an OEM relationship, and instead went internal in UFS 4.0. All of SIMO's UFS sales to Micron didn't transition to UFS 4.0 for this reason. These massive speedbumps every three or four years killed stock performance while they were worked through. So........now we have the NAND makers exiting UFS cell phone controllers, and the entities that are replacing them are not SIMO competitors, they are SIMO customers. SIMO is not going to lose cell phone controller business to the NAND makers internal teams, they need to worry about other UFS controller makers. Fortunately, there are none of any meaningful size. There are teeny tiny competitors (Maxio) which are like Gateway compared to Dell. Pesky bugs. SIMO being a market share driven price sensitive Asian competitor will kill those guys. So.....the main difference with SIMO now and in the past 15 years is a huge chunk of business is not going to be lost to a NAND maker and a year or two spent regaining share with another OEM. Sales are only going to go up.
What else?
Autos. This is going to be 10% of SIMO's sales in 2026. It's growing rapidly RIGHT NOW as design wins from 2-3 years ago enter auto production RIGHT NOW. SIMO sells some auto modules which include a controller and procured NAND memory in their Ferri product line. They also have an OEM relationship with Sandisk for UFS controllers in autos. Sandisk uses a SIMO UFS controller, SNDK NAND memory, and packages it into a module made for use in a car, and Sandisk markets it to the auto industry. They also have an OEM relationship with Samsung for eMMC controllers in autos. Samsung uses a SIMO eMM controller, Samsung NAND memory, and packages it into a module made for use in a car, and Sandisk markets it to the auto industry. NAND memory controllers in autos are growing rapidly (EV or not) as cars become more like computers on wheels, have video programming options, and this is just a growth segment for at least a decade. SIMO's dominant position in eMMC and strong position in UFS means they are highly likely to participate in this auto industry controller growth. It is less "locked in" that cell phones and PCs, but it seems like a great business, for years.
The two sexy MEGA new areas:
- Enterprise controllers. SIMO has a new product line (Mon Titan) that is a controller for data centers. It started sales in Q4 2025, and is supposed to do $50m in H2 2026. If it's successful, the sky is the limit. They are spending a shit ton of money on R&D for this product line, and historically when SIMO enters a market they own it in about 3 years. The Chinese NAND maker Yangste is an OEM for this Mon Titan chip, meaning enterprise SSDs sold by Yangtse to Chinese AI data centers (or anywhere) likely include a Mon Titan controller chip. The volume potential is massive. And they are taping out a state of the art 4nm enterprise controller chip (Mon Titan is 6nm) in Q2 2026, and that product has two NAND makers (presumable Yangste and one more) as OEM customers. If the NAND maker is successful with one of these products, the volume potential is huge. It's not hard to see Mon Titan doing $50m in H2 2026 and then $200m in 2027 as the sales programs ramp toward maturity. And then, after that, they continue to grow because TLC and QLC SSDs are the growth area of AI data centers (Mon Titan is the controller for a TLC or QLC enterprise SSD). This is a great segment to pursue (if successful), but it's a 2027 and beyond story for REAL sales. Fortunately stock prices like to have something to look forward to.
- AI data center boot drives. What's that? Well, I'm not going to pretend I completely understand AI systems architecture, but SIMO is in NVDA's Bluefield 3 system with 1 SIMO boot drive per system, and seems likely to win 4+ slots in the 2026 Vera Rubin system at 3x the price of the Bluefield 3 boot drives. In other words, sales of this product are going to explode higher (starting from zero last summer). Another interesting (and complicated) thing about boot drive is.......SIMO has to procude NAND memory, package it with a SIMO controller, and sell the entire finished boot drive to the customer. The controller is perhaps $100/each, but the NAND may be as much as $5,000 per drive (not really sure). So.....if this segment takes off, it's going to be HUGE (SIMO's PC SSD controllers are about $6 per chip), but with low gross margins as SIMO can't just mark up the NAND portion. But, at $5,000 per boot drive, it may be just stinking massive. This also is a 2027 story before sales become really meaningful, and stocks like something to look forward to. Google is also a direct customer for boot drives.
Both of these two sexy new MEGA segments will benefit from an environment of lower priced NAND memory chips. As you all know all memory is currently in excess demand, primarily HBM and DRAM, and surprisingly NAND as well. Only Micron, Samsung and SK Hynix make HBM and DRAM in size. HBM and DRAM are more profitable than NAND, so those three are expanding their HBM and DRAM capacity much more rapidly than they are expanding their NAND capacity (all three of them make all three types of memory). Fortunately in NAND there are three other major global companies that make only NAND (Sandisk, Japanese Kioxia and Chinese Yangste), and those three are expanding their NAND capacity as quickly as possible because it's stinking profitable as hell. When they expand capacity, NAND prices will (hopefully) decline, and SIMO's ability to procure NAND for AI data center boot drives will improve. And for Mon Titan greater availability of NAND is described by SIMO as a main bottleneck for the growth of this product line. It means that when more NAND comes on line (as it will) the two sexy MEGA areas will be more likely to see the explosive growth that is their potential.
In summary, legacy stuff (90% of 2025 sales - gadgets, cell phones, PCs) is going to grow well for 2-5 years due to market share gains. OEM relationships in these areas which historically caused massive speedbumps in SIMO's sales are no longer a problem as the OEMs are increasingly exiting the segment (and causing SIMO's market share gains). NAND controllers in autos (10% of 2026 sales) are just starting to ramp now, and should grow for at least ten years. And the two sexy segments (enterprise flash controllers and AI data center boot drives) have begun sales late in 2025, and their massive success (or lack thereof) will become apparent in 2027. So....legacy stuff grows in 2026 (and beyond). Autos grows forever. And new sexy MEGA stuff grows (hopefully) in 2027. NOTHING contracts!
I think buying SIMO now and not selling and just holding it through at least end of 2027 is probably a great stinking investment. Even if the AI trade collapses, SIMO's share price will drop with the group, but the SIMO story is fine - 90% of 2025 sales in gadgets, cell phones and PCs will grow due to the market share gains. More likely the market share gains in consumer products occur, auto growth goes on for a long time, and if 50% of the sexy MEGA segments work out, it's fine.
What else? I don't have a price target or anything like that, it's more like be invested in a stock that is going to (non-cyclically) grow sales.....a lot.
SIMO remains an acquisition story. The larger their sales get, the more desirable they are for a larger semiconductor company to acquire (worth the time and effort). Also, the larger the sales become, the more clear it is that their consumer product segments are not going to be lost to competitors, rather SIMO will be the "monopoly" company in the segment which smaller companies try (and fail) to dethrone.
Interesting recent bits of SIMO news.
- Inventory in Q2 2025 was about $240m. Inventory in Q4 2025 was about $420m. That inventory growth is to support the coming sales growth
- SIMO says their Q1 2026 sales are going up sequentially mainly due to market share gains in cell phone controllers. QCOM says their Q1 2026 sales are going down more than normal seasonality due to weakness primarily in Chinese handset makers (SIMO's main cell phone area). This is evidence of SIMO's big market share gains in cell phones. When the cell phone segment "bounces back", SIMO's sales will get another bump up from the bounce back (they're right now growing sales in a declining unit market).
- MXL and SIMO are in Singapore arbitration over the cancelled merger. SIMO will probably win, and the settlement amount is probably $200m minimum and sky is the limit maximum. That announcement could come anytime between April and September 2026.
Ok, there it is. I guess I'll reassess the SIMO story when it's $200, or when it's Q4 2027, whichever comes first. But if snippets of success in either of the two sexy MEGA areas begin, SIMO can easily go to the sky in valuation.
What's better than that in tech? |
