| | | Notes from SIMO investor presentation from two weeks ago (March 17 2025)
Q1 demand comments – there have been pockets of strength in Chinese cell phones, and 4 NAND makers have recently talked about increasing prices by 10%. SIMO previously expected such price increases to occur in Q2, so this earlier announcement indicates demand may be “stabilizing”.
SIMO still expects Q1 to be the worst Q of 2025 and there should then be sequential sales growth through Q4.
He quoted a report that said 82% of enterprise expect to purchase “AI PCs” in 2025. In October 2025 Windows 10 ends customer support.
The high end SIMO PC SSD controller has design wins with 4 of the 6 NAND makers, and they expect to have more than 50% market share of high end PCs.
SIMO’s UFS 4.0 solution comes out this summer, and they have 1 NAND maker OEM and several module makers that begin sales in H2 2025 and next year.
They have a major cell phone maker which purchased QLC UFS controllers directly from SIMO last year for one phone model, and it went well. This cell phone maker is talking about expanding direct controller purchases from SIMO to more models in 2025, and that won’t appear in the June quarter but it should appear in the second half of 2025. To the extent this happens (direct controller sales to cell phone makers) the loss of Micron and other NAND makers as OEM partners is lessened.
For Mon Titan the target customers are large hyperscaler customers, large enterprises (directly), large server makers (Dell, Lenovo, etc.), specialized system companies like VMWAre or Pure Storare, and/or large module makers. SIMO expects their Mon Titan customers to have multiple customers of their own. Their two Mon Titan first customers received Mon Titan product in H2 2024, and they expect those customer to begin shipping to their customers in H2 2025. So H2 2025 is the real start of Mon Titan material contribution to SIMO sales.
The whole Mon Titan project will ramp to 5%-10% of sales in 2026 or 2027 timeframe. That’s the first target, and is not a peak target.
Why is Mon Titan unique? Historically in the enterprise custom ASIC chips – especially in AI servers - have been the controllers, and they work with SLC and MLC NAND memory. Mon Titan works with those types of NAND, but it also works with TLC and QLC NAND which is much less expensive than SLC and MLC NAND types. SIMO thinks AI servers will transition to QLC NAND for memory cost savings if possible, and Mon Titan works great with QLC NAND. If that market share shift (from SLC and MLC NAND to QLC NAND) in AI servers actually occurs Mon Titan will benefit greatly.
As an example, he said Mon Titan handles 128TB QLC memory per controller. The leading controller competitor today handles 16GB of SLC or MLC memory.
ASP guidance – Regular PC SSD controller = $5-$7, new PC SSD high end controller = $10-$15, Mon Titan = $40-$70, mobile controllers = $2.50.
How’s the quarterly revenue ramp for 2025 look? The main driver will be the new high end PC SSD gen 5 controller. PICe gen 5 will contribute somewhat in the June quarter, and moreso in Q3 and beyond. Second driver is UFS 4.0 (starts in the summer). Then Mon Titan which will begin to contribute in H2 2025 with sales and improved gross margins. They plan to give additional guidance about Mon Titan each call, and when they give Mon Titan updates they hope it gives listeners greater confidence in the second half of 2025.
Can you exceed historical 48%-50% gross margins? We expect to get toward the high end of that by the end of the year. Mon Titan and UFS 4.0 give confidence that SIMO should stay near the top end of that range going forward (ie, 2026) and potentially move higher if Mon Titan and UFS 4.0 do well.
MXL Singapore arbitration begins in October 2025 and they expect it to be resolved in November 2025. |
|
