Last week's post on: Why is MediaTek So Damn Good? which was written after it was announced that MediaTek grew 21.7% last year, produced some extremely interesting responses.
'Cheese' listed three reasons for MediaTek's success:
First: For MTK, "value" is how their customers define it. Although this sounds obvious, check with the western companies and you will hear something else. ("We the chip companies are experts, our customers are morons - what do they know about this high-tech industry"?)
Second: MTK is customer-focused. Again, sounds obvious, but MTK really knows what customer focus is. A field application engineer of MTK is not a bimbo - (s)he is versatile, adept at hacking any part of the system (literally overnight) to delight customers. No long technical discussions, no commercial intervention, no project rescheduling, no complex organization behind the engineer -- plain simple "just do it" mantra at work.
Third: Work-ethic. Engineers are not obliged to follow (seemingly western) "engineering rules" - no documents to be made, no "permissions" to be taken from senior architects or technology managers, no lofty ideals to abide to. Then there is no excuse for not slogging. Only results matter - and by results, they mean "business" - not a research paper, not a technology demo, but "customer contracts" or "orders". MTK is (to a large extent) a one-stop shop. Making chips alone doesn't satisfy them. They offer a higher value-proposition, by bundling turnkey software offering along with the chips. They don't shirk away from this responsibility by pointing towards third-party software vendors. When they do work with ISVs, they manage the ISVs themselves, thus keeping a single point of contact for their customers.
To summarize, MTK is to a large extent, a "no bullshit" company, where bullshit is anything that doesn't directly generate money for them, here and now.
The second comment, from Robert, contrasts MediaTek's attitude to customers compared to some Western companies..
Robert writes: TI's Locosto chips were made for the Chinese market, they were cheaper and equally, or better, spec'ed "as phones" than MTK's product.
BUT there were two big problems with the strategy - Locosto was targeted at a simple cell phone only market (not the product that
'Here's a chip with a half done reference design, and some half finished documentation but no support. Just employ 200 -300 engineers and you can probably make a product out of it, (it shouldn't take you more than a year).'
ST and Infineon were only a little better.
What I find more interesting is that there are a couple of companies taking sockets from MTK, so these are the guys to really watch, especially when these new guys are very profitable.
A third comment came from Arun Demeure:
One thing that's especially amazing about MediaTek is that their gross margins are excellent, especially in wireless (although under a bit of pressure). This is quite amazing for a value supplier with no clear technological advantage in a competitive market, and proves that their great reference designs and field application engineers are more than just about grabbing design wins for the sake of market share.
Subscribe to this blog's feed
Comments (6)
The biggest difference between Mediatek and companies such as Infineon and ST-Ericsson is that Mediatek is a software company NOT a hardware company. The do manufacture the hardware but as your article stated, it is not best in class. The real value of the Mediatek solution is the software and reference design which significantly reduced the R&D investment required by the handset manufacturer.
Posted by Anonymous | December 1, 2009 3:15 PM
Posted on December 1, 2009 15:15
Thanks Anonymous, that's very interesting. I suppose that's a smart strategy with the software cost increasing as a proportion of the total cost.
Posted by David Manners
|
December 1, 2009 3:23 PM
Posted on December 1, 2009 15:23
I would suggest that MediaTek being a Taiwanese company with close ties, culturally and business-wise in China would have recognized that the shanzhai players needed a decent software stack on a semi-acceptable and cheap platform.
MediaTek was probably more aware of what the local Chinese market could accept in terms of quality and experience.
If it looks like an iPhone, has 40% of the iPhone functionality then there is a huge market for a much cheaper Chi-Fauxne :)
The Chinese want their brands and culturally its no stigma to support your favorite brand (clothes, cars, phones... whatever) with a knock off. It still says... "I like Apple"
Tai-Pan
www.shanzai.com
Posted by Tai-Pan | December 2, 2009 7:04 AM
Posted on December 2, 2009 07:04
Well yes,Tai-Pan, but surely the likes of TI, Qualcomm, ST and Infineon have the nous to figure that out for themselves. Or am I over-estimating them?
Posted by David Manners
|
December 2, 2009 7:38 AM
Posted on December 2, 2009 07:38
David,
I have to agree with Tai-pan, TI and the other big boys simply didn't have a clue how to deal with the shanzai, and were only marginally successful with ZTE and Huawie.
At last estimate there are about 200, that's right two hundred companies in China trying to do cell phones. Most of these guys were making MP3 players last year and have little or no idea how to make a cell phone. So almost ALL the technology comes from Mediatek.
TI or ST expects the phone maker will behave like Nokia, Motorola or Sony/Ericson and will want to port their software to your phone hardware, so they can run their stack and HDI (to differentiate their product)
Consequently TI only provides the basic functionality in a reference design. Mediatek reference designs however, are production ready, and field proven / debugged and cost minimized solutions. Most shanzai use Mediatek reference designs "as-is" the ONLY change is the start-up screen where they add their logo, or more often Nokia's logo.
There are other things like multimedia, bluetooth file transfer (song/video swapping), and touch screens (with Chinese character recognition) all essential entry level phones in China.
From what I heard, TI didn't even support BT song swapping, I guess they are concerned with RIAA/MPAA. TI also wanted to sell OMAP/Davinci as a $10 multimedia add on to Locosto. They wanted to keep the multimedia phone differentiated from the basic phone, and bye the way figuring out how to incorporate a touch screen controller was your problem.
Anyway, as I have said before, what you are seeing is part of the big revenue transfer from the old guard to the new entrants, with a new business model. NXP's home business unit was made worthless by Mediatek's big wins in the HDTV space (hence the sale to Trident for pennies). I'm guessing the ST/NXP/Ericson GSM venture is also not worth what ST paid NXP for it.
Posted by Robert | December 2, 2009 10:07 AM
Posted on December 2, 2009 10:07
Robert there's some amazing stuff in there - 200 companies making cellphones! Thanks that's a really great insight.
Posted by David Manners
|
December 2, 2009 11:01 AM
Posted on December 2, 2009 11:01