| | | Data centers represent a lot of growth
SIMO doesn't sell into data centers. They are trying to enter the market now, and their success there will become known in 2026 or 2027.
a lot of data centers are using consumer-grade controllers.
Really? How many "consumer-grade controllers" were sold into data centers in 2023? Who sold them?
IoT represents plenty of growth for their eMMC segment, which before the inventory correction issue was growing very nicely, and presumably will return to nice growth again.
Really? They don't break out eMMC sales separately from UFS sales. How do you know that eMMC was growing nicely, and how do you have any idea how large this segment is?
An eMMC controller used to cost about 25 cents. They probably cost less today. With that price per unit it's hard for this segment to be material.
Auto is not a particularly new industry for them, and they've been growing in auto and industry for a while.
Really? My understanding is auto shrank in 2023 and 2024, and may resume growth in 2025. Growth from low levels in 2024 isn't hard, but the numbers may be small. If you have any data on SIMO's auto sales, please share it.
I also don't see them as direct competitors to the NAND makers and neither do the NAND makers view them that way.
I think each Mon Titan data center SIMO controller sale will take market share from a NAND maker's data center controller. That's direct competition.
Micron used to buy UFS 3.1 and below controllers from SIMO, and was probably responsible for 80% or more of SIMO's UFS controller sales. From UFS 4.0 Micron will stop selling memory modules with a SIMO UFS controller and instead sell memory modules with a Micron UFS controller. That's direct competition.
I think they are heavily favored to break into and gain good share in the enterprise controller market.
Perhaps. We'll know the answer to this in 2026 or 2027. It is also possible that the NAND makers are happy to give SIMO the high volume, low low end consumer product segment design slots, and are not eager to give SIMO the high end high margin enterprise segment design slots. In other words, perhaps SIMO's fails miserable in enterprise controllers. We really don't know.
SIMO is spending money to succeed in enterprise controllers today. If they fail, the money is gone. If they succeed, I doubt the sales success is strong enough to make their previous target operating model (50% gross margins, 30% operating margins) feasible. So far, all we've seen from SIMO's efforts in enterprise is a significantly higher cost structure. Whether the future sales justify those costs remains unknown.
At the temporarily depressed margin level they are experiencing now, they still have a P/E (ex-cash) of under 19.
When SIMO was growing like a weed from 2010 to 2018 the forward PE was usually about 11x-14x. That was a typical Taiwan tech/semi stock valuation. SIMO's sales were about $28m per quarter in 2010, and now 14 years later SIMO's sales are about $210m per quarter. That's why it seems very expensive to me at 19x. It used to be 12x, and today there is no ongoing transition of previous tech to NAND memory in high volume consumer devices like there was in the previous decade. And, margins area depressed. Everything is worse! Even the tax rate is higher than it was five years ago.
But the stock valuation is higher. It doesn't add up.
SIMO is a consumer product commodity chip stock, without a technology transition (cell phones using NAND, PCs changing from hard disk to NAND) ahead of it that should increase sales by 25% per year for the next 2-4 years. That's why 19x is a demanding valuation.
Blue chip semi stocks have gross margins around 65%. Consumer product commodity chips stocks have gross margins around 45%. They're different.
My hunch is that as SIMO reports quarterly results, the things I'm complaining about will become more and more obvious, and the multiple will contract. Heaven forbid they ever cut guidance, that would cause the multiple to go right back to 10x pretty quickly, which might be about $50 or so. |
|
