|I'm listening to the call now, I'll try to write down important new stuff.|
CEO prepared remarks:
NAND industry supply conditions are the biggest factor limiting SIMO's growth. NAND flash makers are directing available flash to enterprise applications and away from client SSDs, and this trend increased in the past few months. Flash supply is inadequate for client SSD. Supply will improve over the next few quarters, and flash makers will transition from 256GB units to 512GB units.
SIMO expects to ship QLC client SSD controllers in H1 2018.
Flash makers are increasing CapEx, and the tightness since 2016 will reverse over time. NAND supply will reverse and become plentiful for a few years to come. Flash industry cyclicality has not changed, excess supply will come, it is just taking time.
Last year NAND makers took NAND from module makers to their own client SSD products, recently NAND makers have took NAND supply from their own client SSD products to their own enterprise NAND products. SIMO's pipeline is unchanged, the timing of client SSD projects has just been pushed back a bit. Programs will ramp in Q4 2017 and beyond.
We are well placed to benefit when NAND supply improves.
We have added a second customer for enterprise SSD controllers, and now have two hyperscale "direct" customers. One of these programs have been pulled in and will launch before the end of 2017.
Our new UFS controller program remains on track, and we expect initial shipments to a top 5 Android OEM to begin later in 2017. For most cell phone makers we think eMMC will be preferred over UFS now due to price differences.
Shannon and Ferri:
Tight NAND supply limits out growth. If not for NAND supply both Shannon and Ferri would grow meaningfully this year. High grade NAND is available, but very expensive now. Ali Baba (he named the Chinese Shannon customer) continues to purchase Shannon products from us. Ferri is doing well. We have Japanese, European, Chinese and US commercial OEMs, and we have signficant orders we cannot fill because NAND flash is not available.
We have a strong pipeline and are well positioned in all three segments to rebound when more NAND supply comes on line.
CFO - Finances.
Client SSD controller sales we weak because high prices slowed adoption and NAND was directed toward enterprise. Beyond this year we remain confident of rapid growth, most future NAND products are much cheaper than current NAND, and the replacement of disk by SSDs will accelerate.
eMMC controller sales will decline in H2 as SK allocates flash from mobile to SSDs. Longer term eMMC will be stable (but this sounds like BS to me, I think they're losing SK's UFS biz).
Shannon and Ferri - SSD solutions will grow in Q3, and peak for the year.
Full year client SSD will be flat, eMMC down slightly, and SSD down by less than previous guidance.
Head count is 1,153. Tax was 22% compared to 27% in Q1. We think the effective tax rate will remain above 18% due to inability to realize tax benefits of entities generating pre-tax losses.
Cash is up $86.9m year on year. Accounts receivable and inventory are both up a lot sequentially.
Revenue will rebound and gross margins will also rebound from Q3 weakness. Q3 projects were pushed back to Q4 due to NAND supply weakness.
eMMC in second half will be down compared to H1. SK Hynix built inventory of eMMC in Q2, and those numbers will decline in H2 as Sk will allocate flash to SSDs rather than mobile.
MRVL and internal solution are showing price competition? No changes to competitive landscape. We dominate the client SATA controller market, PCI controller products are coming on line later.
How can client SSDs sales be flat with Q1 down 20%, Q2 down 10% and Q3 down slightly? No real answer, but Q4 will rebound and client SSDs will be flat for the full year.
How do we know NAND capacity will expand going forward? It hasn't happened yet. Client SSD for SIMO will rebound whether NAND supply improves or not. We have a major PCIe program to launch in Q4 with 3D NAND.
For eMMC they may add a second NAND maker as OEM, and enter new markets other than smartphones.
Q When will supply-demand get more "normal"? A: Flash makers are ramping 3D NAND and spending Cap Ex to ramp capacity sooner. NAND makers are saying yield and output are ahead of expectation. This year the NAND growth is back half loaded, but 2018 should be a meaningful year for NAND supply growth. So early in 2018 or mid-2018 NAND supply should be more "normal".
10% of SIMO eMMC today goes into non-smartphone applications.
Ferri SSDs have a diverse customer base, Shannon is more concentrated at Ali Baba.
Do you plan to spend all $200m on the buyback? No real answer.
If you executed the entire buyback, are there any acquisition candidates for you? No answer, of course.
Charlie Chan of Morgan Stanley! How about SIMO's SSD growth versus the industry growth? SSD units are increasing even though supply is tight. Last year it was 40% and this year 45%. So....why isn't SIMO growing since the SSD client controller space is growing? Answer - we don't think we're losing market share. SSD prices have gone up a lot, and adoption is slowing. 64 layer and 96 layer NAND should push SSD prices down, and we think that will cause SSD adoption to accelerate. We don't think adoption is increasing in 2017.....
Charlie Chan of MS on Op Ex - Is Op Ex for SIMO in Q3 going down? A: It should be stable in Q3 compared to Q2. Enterprise SSD controllers are contributing to OpEx now, but not yet shipping.
Charlie Chan of MS asks.....will NAND supply increase soon? You were wrong earlier in the year when you said Q3 or so, why should we trust your view now? A: We use the rolling guidance from our customers, and they have been inching higher especially in Q4.
The proportion of Ali Baba where SIMO buys the flash is higher this year than last year.
Again, client industry growth is about 10% to 15% in 2017. But SIMO client SSDs are expected to be flat. Why? A: We think your industry forecast is wrong.
Which products have the strongest visibility for Q4, which is most risky? We are confident across the board for our key products - client SSDs, eMMC and SSD solutions.
Q4 gross margins will snap back due to material rebound in high gross margin client SSD solutions and decline in Ali Baba 2017 program in Shannon Systems. And in Q1 2018 we can renegotiate pricing based on higher prices.
Why is Q3 declining sequentially? Tight NAND and their customers are rebalancing to non-SIMO products.