|MU announced an internal controller for their latest NVME SSD models. That's bad, it means MU may go internal going forward. More likely they will use SIMO controllers for some models and internal controllers for others, and it will then depend on how each model does in the market.|
But pretty bad news all around. I get the impression SIMO should do great in the next round of client SSD growth, and then after that it will have more competition from internal controllers from the NAND makers. Offsetting the MU and WDC trend to go internal is MRVL's decision to exit client SSD controllers completely. I get the impression that SIMO kicked MRVL's butt in client SSD controllers, so they exited the market, now we've gotta see how the internal teams do against SIMO. While a controller may not be that super difficult to make, I think SIMO's appeal is that they have the controllers ready for usage BEFORE the supply chain is ready to make design wins, so with SIMO you get a functional controller (hopefully a great one, but who knows) on time, while with the NAND makers internal teams it may he harder to meet deadline and such. SIMO has millions of R&D dollars to throw at making controllers, the internal teams must be limited on how much they can spend and still be more cost effective than just buying from SIMO.
It's all market share. SIMO can amortize one controller across 20 customers, the internal teams have to amortize their internal program against one product line. If that line just does so so, the internal team could be on the chopping block in the next downturn.