| | | SIMO thoughts.
Well, perhaps the re-rating of SIMO's PE multiple is underway.
The company says that in about a year auto related controller chips (and modules) will be more than 10% of sales, and expected to grow for the foreseeable future. Mon Titan is expected to be 5%-10% of sales in about a year, and if successful expected to grow strongly for the foreseeable future. SIMO will begin shipments into the NVDA eco-system in Q4 2025, and they're targetting more products to get involved with NVDA and their AI driven growth.
In the legacy client device segments (PCs and cell phones), well, in PC SSD controllers SIMO already has the design wins which (they say) guarantee them global market share growth from today's 30% to about 40% in 2-3 years. In mobile device controllers SIMO indicates the flash makers are de-emphasizing modules and the trend is toward discrete purchase of controllers and memory by module makers, or by cell phone makers themselves, and then the module maker or cell phone maker assembles the module (memory + controller) themselves. MU has already indicated it is exiting mobile managed NAND, so new internally developed Micron flash controllers will no longer be a competitor for SIMO. SK and Samsung have (I think) said they generally intend to de-emphasize NAND and will instead focus on the more profitable HBM and DRAM segments. One way to "de-empasize NAND" is to not spend the resources to develop new internal NAND controllers and instead buy them from companies like SIMO. In eMMC SIMO has the most complete product set already on the market.
All that adds up to visibility and growth in the cash cow legacy client device segments (PC, gadgets and mobile) and reliable strong growth in the new segments (auto, enterprise and NVDA eco-system) which - together - perhaps justifies a permanently HIGHER PE multiple. SIMO traded at a PE of about 10x to 15x for most of the past 10-15 years, maybe their PE is now more appropriate at the 20x-25x level? Why not? Diversified semis trade at those PE levels, and few semis have the growth characteristics and market dominance which it seems SIMO now owns.
Another point - SIMO says they are the only flash controller maker which does business with EVERY NAND maker. In order to make the merchant controller work properly with the NAND maker's memory there probably needs to be some sharing of roadmap and NAND technology between the memory maker and merchant controller maker. The memory maker doesn't want to share this (perhaps sensitive) information with lots of companies - in fact, the fewer the better for the NAND maker in their efforts to preserve their intellectual property. But, in order to have controllers ready which function with their future NAND memory the NAND maker needs to share their plans and processes with some merchant controller makers, and SIMO seems to have that role locked in with all the NAND makers. The point is new entrants in the space which may want to compete with SIMO may find it difficult to develop the necessary trust and relationship with the NAND maker, and it may be hard for competition to break into this controller business with the assitance of the NAND makers. If that's correct (it probably is) it also supports a higher multiple for SIMO as it reduces competition from other merchant controller makers. |
|
