That Barron writer is never very good. TSMC's comment was more specific - weakness due to lower snapdragon orders. TSMC' took away the iPhone order from Samsung so Samsung is paying back by using its own AP in galaxy phones. This is actually good for simo.
Yesterday TSMC forecasted 0-5% Q/Qvrev growth. TSMC Q1 has been very strong due to iPhone 6 chip orders so this forecast was considered to be quite good. It is also a good indication of General mkt demand.
Still the size of the drop seems silly: 1. Q1 should not have much TLC in it anyway, and simo's forecast of 2015 eMMC controller growth can easily be achieved by the SK Hynix business. 2. There is not much of a demand issue, as Digitime reported just a couple of days again that D/S is coming to balance in Q2, and appears to go tight by 2H, based on sources of a major Asian producer (for which I suspect is SK Hynix). 3. PC demand is soft, but on the small business side. Just by looking at how many models are coming out with SSDs, the growth there seems to be intact. 4. Comparing the multiples of those high growth high multiple players, simo is cheap. 5. Look at the volume, institutions are not doing the selling.
As some of the fund managers were calling this as opportunities for some individual stock today on CNBC, I think this is one of them. Anyone agree?
I have pointed this out for months. But Samsung has not moved fast enough in this area, probably because all energies are on the AP. Getting the Carriers to validate also seems difficult for Samsung, and it is strange that the simo transceiver would not include CDMA support for US market, which is sort of half hearted effort..
I don't care Q1 numbers to involve LTE revenue, but whether a Samsung LTE modem will make it beyond Korea because such a move will demonstrate that in-house modem is part of the new Samsung strategy, and we all know what that will do to SIMO share prices.
We know for sure the US version will have a QCOM modem in it to support CDMA. But S6 will also be on sale in UK, India and many other counties on 4/10 as well. Since Samsung launched Exynos 300 series modems last summer, Galaxy Alpha and Note 4 for international markets have been using an Intel modem while Korea got the Exynos modem. Intel has not declared any Samsung design wins lately so it will be interesting to find out what the international version will use.
That happens, but usually with bad news from the sector. We do not have any of that today either. On the other hand, the volume does not spell trouble to me at all so maybe we just have one of those market "clueless" days.
Yes - it is very good to see how they are impressed by SIMO's penetration into SK Hynix. They were very good back in 2013, figuring out SIMO was going to lose the LTE sales before a lot of others did so I expect they get it right this time again but their reference throws me off. To me the transceiver business is just nice to have too unless it becomes part of Samsung's future strategy (with SIMO in it). I do not put a lot of hope on that because Samsung is known to be cheap and if modem is in their future they will try to make everything in-house.
SIMO announced transceiver design wins in Q1 and Q2 2014. Those dates fit Samsung announced Modem 300, which is a cat-4, nicely. In Q3, SIMO said its design wins are in production - Galaxy Alpha, Galaxy Note 4 and Galaxy Note 4 Edge. Those models sold in Korea uses Exynos 300 and 303 (a cat-6 modem). In other words, SIMO are in the latest two known Samsung LTE-A modems. It just happens that S6 also has sports a cat-6 modem.
The first Samsung modems were CMC220 and 221 in 2011, but they were for FDD-LTE data only. Samsung added WCDMA/HSPA+ and GSM supports in 2013, but only for Korea, no flagship models. Samsung used QCOM MSM8974 in all its LTE models in 2013 because that modem was the only one to support LTE-A in 2013. QCOM had 95% of market share worldwide that year.
In 2014, Samsung started to use QCOM, Intel and its own, but the Intel modem was used for many of its international versions. The Intel modem was launched way ahead of Exynos 300, giving the impression that Exynos 300 came too late for both Samsung and SIMO. Samsung used complete solutions from Marvel as well in 2014, but only for low price models.
In summary, there are only 2 known Samsung LTE-A modems right now, and there are evidence that both use SIMO transceiver. There is no evidence of a Samsung transceiver exists!
There are only 3 known LTE-A modem suppliers that Samsung has used for its flagships (QCOM, Intel and its own), and all 3 have cat-6 modems. QCOM is the only one that supports CDMA (SIMO transceiver does not have CDMA) so it is safe to say that S6 US version will have a QCOM chip, but for international market, the answer is any one of 3. SIMO transceiver may not be in S6, but at least it will not because Samsung has an internal transceiver.
There were higher trading volumes on 2/25, 2/26 and 2/27. So today's reaction is probably by small investors who did not have access to that report back then.
But even those higher volumes are not high enough to put any meaningful dent to the institutional holdings on this stock so it looks like those big guys have not dumped this due to that report. Needham was pretty good back in 2013 when it was among the first to detect that the Samsung + Simo modem+transceiver lacked required carrier aggregation. So I am surprised that it would make a wrong reference this time around in its latest comment.
Interesting, but its reference to "Sam using an internal solution in 2013" is confusing because there is no 2013 product information to show that. The 2013 models had an Intel solution or a QCOM solution (both are modem + transceiver).
That does not mean that Sam have not developed its internal transceiver since 2013, but at least that reference seems to be incorrect. I also wonder why Wallace would not mention in the conference call a Samsung internal transceiver as a potential barrier for future wins if the product exists. Finally, if Tsai leaked something to Needham that it had not previously made public, I wonder if SIMO has broken some public disclosure rules...
No one knows yet, and I do not think the pricing move reflect real knowledge. It was oversold last week for no reason (except day traders changing minds). On the other hand, I would still bet a quarter that SIMO will be in some version of S6 for the following reasons:
1. It appears S6 will be all Exynos. Getting rid of QCOM means the door is open.
2. Last year was Exynos + Intel modem for international (EU) markets, but hard to believe it is a LT strategy.
3. There were analysts comments Samsung LTE modem getting wider carrier acceptance (not in US yet?).
4. S6 has a cat-6 modem, and Samsung 303 is a cat-6 modem that introduced 5-6 months ago.
In any case, more analysts now believe Samsung's strategy is to sell Exynos to other phone makers. To do that, Samsung cannot rely on a competitor to supply modem (so they can make it difficult). So Samsung has no choice but to make its own, and no choice but to get carrier validations for its modems. With Exynos in S6, I think chapter-2 of Samsung mobile strategy has started, and it is possible by 2016 SIMO transceiver will no longer be a 1-customer business. How will analysts value SIMO If the transceiver becomes the eMMC controller type of business (one partner, multiple customers)?
That article shows the authors does not know that he is talking about. You can read the specifications of Exynos 303 on Samsung's website and find it is 3G and 2G compatible. The question is not about technology, but carrier validation.
Agree - it is a surprise that the rumored Cat-9 is a no show here, but the Exynos Modem 303 is a cat-6 first appeared last September. But QCOM also has cat-6 available. On the bright side, 303 has born a little longer so hopefully it has passed more carrier validations.
Using cat-6 may also allow Samsung to more easily change the modem for different market (like US). In the end, my earlier guess may still be correct that we will see a Samsung modem for international market and a QCOM one for US market. Hopefully SIMO says something soon...
1. I think Samsung's strategy is clear but the launch with Exynos in it will validate that. After that, Samsung may need some sales success to get phone makers to buy Exynos. The first such deal will convince ten analysts and get them talking.
2. Whether s6 for US has in-house modem depends on whether Samsung has passed US validation. Back 2013, Samsung did not have carrier aggregation but that is past now. Samsung has a new cat 9 modem in its updated Note 4 just released for Korea. I wonder why that one is not in US. One possible answer is US validation is not completed yet, and it is possible because ATT makes it super difficult. So this year it is not tech issue anymore. I suspect last year simo design wins were associated with the 303 cat 6 and the 333 cat 9 modems so if in house component and selling solutions is the new strategy it will be just matter of time before people realize simo transceiver will be big again. This also makes sense why simo keeps pouring money in this pit.
To add, current price is below s6 rumor started so from risk reward prospective why are people selling?
S6 wil use Exynos AP because it is part of Samsung's strategy to become competitor of QCOM and eventually dominate mobile AP. If s6 becomes a hit, other smartphone makers, especially those in China will buy the chip for their models.
But to be a real competitor, Samsung has to also integrate modem into AP. This is why Samsung goes through all the trouble to get carrier validations. Now, if above is true, the simo transceiver revival seems to be inevitable, and SIMO will win more designs this year starting with s6.