REGISTER | LOGIN
Breaking News
News & Analysis

MediaTek Faces Eroded Smartphone Growth

8/19/2015 08:54 AM EDT
4 comments
NO RATINGS
1 saves
More Related Links
View Comments: Newest First | Oldest First | Threaded View
sinrohu
User Rank
Rookie
Totally agree with your comments
sinrohu   8/21/2015 10:20:18 PM
NO RATINGS
Totally agree with your comments below.

It will be mostly about their choices for the next 1-2 years if they do well or not. What cores they use on what process, timing and how they execute.

 

Besides, I think snapgragon 620 should be very important for QCOM. If 820 did not show up in time, 620 will be the weapon against MTK X20 for QCOM.

realjjj
User Rank
CEO
Re: eh...
realjjj   8/20/2015 11:15:49 AM
NO RATINGS
I don't speak any Chinese so it's pretty hard to get informed through Google Translate , the US and European press covers Asia rather poorly so can't claim i know enough about Spreadtrum. As far as i know Spreadtrum's top chip is A7 5-mode LTE and WCDMA SoC,  was announced as shipping in April so still low  end. They do have some wins at Samsung but Samsung now has it's own 8xA53 chip (Exynos 7580) so next year they might cover even the low end with their own SoCs. Huawei has a low end tablet with the SC7731G.

Not so long ago here on EE Times Spreadtrum was claiming that they'll go 16ff this year for an 8 cores SoC and 14nm Intel in 2016 in both low end and high end. We'll see how that goes. I did notice that Mediatek is in a rush to go 16ff. They said high end in the first half of 2016 and in the second half the mainstream goes 16ff. It appears to be a reaction to what their competitors are planning, so either Qualcomm or Spreadtrum, or both.

About one week ago Tsinghua's chairman was quoted as saying that they plan to invest 30 billion CNY (4.7B USD) in the next few years to catch up in mobile so Spreadtrum is not giving up.

Something else i want to add about China. First China Mobile reported today and only 23.5% of users are on LTE. Second , everybody is saying that the market is saturated but i would say that urban China is, rural not at all. Rural China is still some 600 million people. In many ways it could be compared to India with a huge difference, data prices. In India data can be up to 10 times cheaper. Why bother buying a smartphone if all you can afford is a 110MB China Mobile package for 2.5$. If data was 10 times cheaper, things could change.

In the end the smartphone market is getting a bit stuck and the next big boom can only come when foldable screens arrive with decent implementations at reasonable prices. Sure PCs and tabs would get hurt but all in all there should be significant upside for anyone that's not mostly PC.
.

junko.yoshida
User Rank
Author
Re: eh...
junko.yoshida   8/20/2015 7:40:18 AM
NO RATINGS
GREAT! Thank you so much for your detailed analysis here.

I think you captured and summerized the mobile chip market in China so well in the following graph:

Mediatek is yet to actually aim for the high end, Spreadtrum is yet to go above low end so the risk is bigger for Qualcomm. The iphone win can always go away and big traditional Qualcomm customers are all declining too.


It's so true, therein the struggle by MTK (and by Qualcomm).

realjjj, you didn't discuss much about Spreadtrum. Where is Spreadtrum at now, then? Are they still stuck in low-end, no brand phones?

realjjj
User Rank
CEO
eh...
realjjj   8/19/2015 12:27:05 PM
"Chinese rivals such as Spreadtrum Communications and Huawei are eroding that position on strong price competition"  - Huawei is yet to manage to sell it's chip to anyone else and their chips are not performing well enough to compete yet. As for the Redmi 2A it sold 5.1 million units in some 3.5 months according to Xiaomi but that's not all that great, it was priced too high at first.

Anyway, his reasoning is valid but he needs to go deeper and that may change the conclusion. The DIY craze will work for some, won't work for others. Apple doesn't have a modem , Samsung is yet to integrate the modem in high end , others are even further behind. It is yet to fully work for anyone  and it will be quite some time before we see how this trend goes for each player.

For the general market and the slowing growth he takes Qualcomm out of it for no reason at all. Qualcomm , Mediatek and  Spreadtrum will compete for the available market. Marvell, Leadcore, Intel will try to gain relevance. Apple, Samsung, Huawei, LG, Xiaomi , maybe Lenovo, maybe Coolpad (there is a rumor in China about Qihoo 360 being in talks to buy some VIA assets but the talks could have been about the sale of a HTC factory) will do what they do and ,at least in part, use their own SoCs. Mediatek has no share or little share at many of those while sales for most of those are also declining.Mediatek is yet to actually aim for the high end, Spreadtrum is yet to go above low end so the risk is bigger for Qualcomm. The iphone win can always go away and big traditional Qualcomm customers are all declining too.

Separating 3G and 4G is somewhat pointless. The modem is not that important. India has 4G and even in China 4G has minimal penetration. China Telecom today reported some numbers and only about 15% of customers are on 4G. Mediatek isn't giving up on the lower end market either, the MT6580 and MT6570 seem to have started shipping - quad and dual A7 with 3G. The loss of share in 3G was as the market moved away from it. You have to factor in ASP too. Loss in share in 2-5$ SoCs vs gain in 10-15$ with room to go for the 35-50$ ASP segment.They could grow on share gain from Qualcomm and mix, with price erosion and size of the TAM as headwinds. The Xiaomi Redmi Note 2 win is very important short term for Mediatek. Likely the best smartphone in the world now, specs per price and it's bound to sell very well. Xiaomi was all Qualcomm and Leadcore in phones for some months now. The Redmi Note 2 sold 800k units in it's first day of sales this Sunday- 6 flash sales every 2 hours with only 2 of the 3 versions available. So just this model can sell 4-5 million units in half of Q3 and 7-8 maybe more in Q4 while using the Helio X10 so high ASP for Mediatek. The Helio X20 in Q1 could be the fastest SoC available if Snapdragon 820 doesn't show up. It will be mostly about their choices for the next 1-2 years if they do well or not. What cores they use on what process, timing and how they execute. Longer term, it might be about the GPU since the CPU perf is hitting good enough. Plus ofc, any innovation that could allow for much lower costs.

PS: The Redmi Note 2 should be the best selling model in China in Q4, if you consider iphone 6s and 6s Plus as 2 distinct models.

Most Recent Comments
resistion
 
resistion
 
resistion
 
resistion
 
R_Colin_Johnson
 
balajis_snps
 
David Ashton
 
traneus
 
traneus
Like Us on Facebook
EE Times on Twitter
EE Times Twitter Feed