SHANGHAI – Let’s face it. All along, we knew how this movie would end. The game is over for Western chip companies--except for a sole survivor and the mobile chip industry leader, Qualcomm.
Earlier this week, upon returning to my hotel from an interview with Spreadtrum CEO Leo Li at his office here, I found in my e-mail inbox the announcement on the breakup of the ST-Ericsson joint venture.
Ericsson is taking over the 4G LTE multi-mode slim modem product line. ST alone will oversee existing products including the legacy modem business, RF, Power Management and NovaThor integrated apps processors, according to the announcement.
A week ago, Renesas Electronics back in Japan announced that it’s “reviewing” the direction of its mobile business. The implication is that it has decided to divest Renesas Mobile or explore alternate business models for its mobile subsidiary.
Both ST-Ericsson and Renesas Mobile, in search of buyers, are likely to ditch their mobile baseband business, just as Texas Instruments, Freescale and Analog Devices did years ago.
Strategy Analytics’ analyst Sravan Kundojjala, in his commentary posted Monday (March 18), blamed ST-Ericsson’s breakup on “duplication among legacy products, transition to a new product roadmap and constant management changes.” The analyst said that the JV struggled to integrate multiple companies and execute on its original plan.
Other industry observers might posit similar reasons for Renesas’ failures, since Renesas Mobile, too, is a company based on the integration of several different companies--including Nokia’s former modem team.
Kundojjala may be correct. But I think his analysis misses a crucial point: both companies failed to recognize how a traditional business model--serving tier-one handset customers--became irrelevant to ultimately winning the global mobile battle.
As Li put it, 10 years ago, “you were OK,” if you got design wins from handset vendors in the United States and in Europe. It’s because “those tier-one guys--Motorola, Nokia, Samsung, Sony-Ericsson, etc.--held 90 percent of the global market share.”
Today, the share of the traditional tier-one companies is only 40 percent. The rest of the market is China, according to Li.
Further, China’s smartphone market today, estimated to be 240 million units in 2013, is much bigger than that of the United States, projected to be about 125 million units, according to market research firm Canalys.
And guess what?
Practically every smartphone–regardless of brand--is manufactured in China today.
Without addressing the burgeoning Chinese market and the needs of local handset vendors who demand chip suppliers to do a lot more hand holding, mobile IC vendors like ST-Ericsson or Renesas Mobile have seen their market share sharply plummet in recent years.
After the leading edge of a technology (e.g., smart phones) passes, volume and revenues favor companies good at manufacturing or who have a few very-high-volume products. But a new wave of technology will favor companies that are good at engineering the new products -- at least for a while, until that edge passes. Companies good at both, like Intel and Samsung, are more likely to do well in both parts of the cycle than companies that are not as deep technically. So what's after smart phones? Wish I knew, but I would look at companies with deep technical benches to do better then, if they survive that long!
And then -- there is an apple.....
The Apple among all the oranges.
And there are national strategic and economic interests of various countries - supporting native sons - among OEMs, IC vendors and foundries
Due to outsourcing and financial "engr." the buying power of the US and W. European customer is going down. For the very same reasons the buying power and technical capability of the Chinese is going up. For cheaper high volume items like Smart Phones the cross-over point is in sight. Soon there will be a mass selling off of related technology from second tier W. corp.s to China. Perhaps ST and Renesas will be the next to cash out.
I don't think we are doing apples-to-apples comparison here. Comparing Qualcomm and Spreadtrum is like comparing an high speed jet with a passenger bus. Who is driving the technology hard? Qualcomm or Spreadtrum? Just saying.
What would be interesting to know is how many of non TD-SCDMA chipsets Spreadtrum actually sells. Its success is largely based on that technology which is used by just one operator in one country and it got lucky.
Agreed that the volumes are huge but why would anyone put all their eggs in one basket? R&D budgets are always tight. STE and Renesas tried and failed to serve tier-one handset makers but where would you rather invest resources in? Low cost, high volume single market or rest of the world? TDSCDMA is not a very attractive proposition for R&D when WCDMA is used by the rest of the world. In any case, western chip companies will not be able to sell at the price Chinese companies can.
Moreover, if you had to choose, who would you rather compete with? Qualcomm or Mediatek? I think Intel has been attacking both and benefits from low cost high volume sales as well as high-end phones.
Thank you Junko -- you can bet that each of them thinks hourly about that same subject.
Plus Huawei, Lenovo, ZTE and several other already are or soon will become "Tier-One" mobile device vendors
Please keep a stream of your wonderful articles coming.....
Well put, and well debated, sranje. No doubt, you could say "the movie we want to see" has just begun.
But the point that I was trying to make in this piece was following.
If a Western chip company doesn't have a "strategy" to serve a growing number of Chinese design houses/OEMs/ODMs,they will see a shrining market. Sure, a growing number of new features and technology advances are to be expected in the next-generation smartphones. But again, what's your strategy to serve "non-tier one" companies?
2. But there is another, perhaps even more important reason. The evolution of smartphones and tablets has not ended -- it has ONLY started. We are just entering an era of accelerating changes, of wireless communication having speed and bandwidth of wired Ethernet and broadband connections. LTE, then LTE-A, and on and on will completely transphorm today's smartphones, tablets, etc. into new devices with nearly unimaginable processing capabilities - 20+ computing cores, new generations of OSs and application SWs that will manage and utilize them, etc., etc. Think voice generation / recognition and processing power required, think of possibilities of new sensors, new non-volatile memories, GaN-on-Silicon power conversion, energy harvesting, new battery technologies, data transparency and data use. Where do you think the future will be forged? Who will be the winners?
In summary - there will be plenty of space for winners in this accelerating transformation of mobile devices for current and future winners - in China and in Western wordls.
The "movie you are writing about" has just started
It is difficult not to like Spreadtrum -- a fast acting, highly enterprenerial, fast growing and successful China's Mobile IC vendor. However, as much as I like your articles, I don't think that you know yet how "this movie will end" -- for at least two reasons:
1. Qualcomm has been hugely successful and is becoming more and more so. But there are also Broadcom - in 2012 probably the fastest mobile processor company on Earth (even if you include China's whitebox tablet processor vendor - Allwinner). And there in nVidia - a company that has proven delivering a stream of very advanced processors in a record time -- unfortunately to date in a limited portion of tablets - squezzed between huge market shares of Apple and Amazon. And there is the Intel - vendor of advanced baseband processors and transceivers. Both Intel and Broadcom are leading suppliers to Samsung (60M smartphones in 4Q12). And then - the Samsung itself - a unique giant who might be on threshold of achieving its decades-old objective to become larger than Intel. Your dirges about Western mobile processor vendors are way premature.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.