Junko is right with "a lack of leadership, or an endemic case of indecisiveness". For instance,"Renesas Mobile, which is looking to become 'the Qualcomm alternative’ in the mobile world" will likely become a dream. As of today, no major mobile phone OEM/ODM ever announced the use of any RMC's SOC. In fact, I am wondering why none of the most recent (of last 10 years at least) leaders of the Japan-based SOC companies ever show any spirit of the Samurai by publicly admitting their shame, given that their companies keep downsized in the past 15 years.
I am not sure if Samurai exist any more except for the Western films...but you have a point here. Definitely the Japanese chip companies are struggling.
How such a proposed consolidation will be translated in the market size, please read Dylan's story here:
A good business is to make money, to take all the necessary action to right the wrongs, to become prosperous, and to ultimately benefit both the investors and the employees (and their families too). Many of these Japanese companies have aging products, products designed specifically for certain customers (particular of Japanese nature because of the language barrier) and most of them are not the so-called world products, no active & intelligent engineering efforts to "localize" the products, boated human resources, and worst of all, management who are not held accountable for their inabilities. Looking back, in the last 10 years, Japanese companies rarely made as much money (percentage wise) in comparison to their counterparts in the western world. There were more red ink than black ink especially given their size. Put it simply, they are stuck with too many "wrong" products and too many "employees" with the out-dated/wrong skillsets. Due to the social infrastructure, they are not allowed to implement massive lay-off. If the rumor turns out to be true, the result is mostly a slower eventual death. Looking back at history, NEC Semiconductors was rated #1 for memory/microprocessor in 1987 and #10 for SOC/Fabless in 2007 (per Gartner/Dataquest). It now exists only under the shadow of Renesas and is about to spun off hence no longer a whole.
Japanese semiconductor companies are unique in the way that they operate like charity houses.
The Japanese are getting old and slow.
Challenged by the combination of fabless companies in the US who have xferred technology, designs and business to Foundries in lower - cost Taiwan and So. Korea ( who are hungry for business even at low margin ) and the consequent loss of market share in the still enormous N. American market, once huge Japanese brands ( like Sony ) have failed to respond ( with technology, cost - cutting & scale ). Instead system and component ( semiconductor ) houses in Japan have gradually abandoned the price driven N. American market, remained content with their own domestic market that allows high margin and consequently have been shrinking.
But Japan can ill afford to abdicate the cost-driven but still enormous US consumer electronics / auto market to the So. Koreans as even low margins generate huge profit which is then inevitably invested into new technologies that ultimately drive established but smaller competitors out of the market.
Japan has double the population of Germany so cannot follow the strategy of the latter of exporting just BMWs and machine tools. Japan has to continue exporting mid-range consumer products as well.
Mega catastrophe like last years Tsunami still brings out the old Bushido spirit.
Just look at how after some initial fumbling they Japan has undertaken a massive clean up & rebuilding that would have daunted any other "mature/western" societies.
Japan is capable of doing the same for semiconductor as well and triumph over the combination between the Fabless wonders in US and TSMC / Samsung. With whatever system / OEM business Japan has left they can afford at least 2 state of the art 28 nm 300/450 mm Fabs / Foundries and keep them loaded.
To pull ahead of the newcomers who get free access to US R&D, Japan needs to invest a lot more in its own R&D and become aggressive in turning the results into innovative end products.
In review of the country's transfer of payments for the past four quarters and the exceedingly anemic economic growth for the past decade, it could be deduced that their Central Bank is literally strangling the country's entire economy, right before our eyes.
Naturally, their semiconductor industry is drained and in full retreat.
I am going to play the devil's advocate here. In the EE Times Forum and everywhere else I look, this proposal, assuming it's really on the table, is greeted as dead on arrival. Like a desperation move. Well, maybe it is.
But my question is: Why not? These companies have operated for years building chips for their own companies' end products or for mostly Japanese customers. It was a good arrangement in its day, but times have changed. Why not pool their resources and see what they can do? Their best option is to get all of their best designers together and see what they can do (kind of like an All Star team, if you will). Maybe it works, maybe it doesn't. But the status quo is not a realistic option. Why not roll the dice?
A good point. In other words, WHAT OTHER options do they have? Right?
If they don't do some sort of a merger like this, they would be just standing still...
But here's the thing. If they are going to get together, and if they are going to pitch it as a "dream team," we do need a leader who can articulate that dream.
When I cover tech industry, what excites me more than anything else is when people I talk to can convey the sense of mission they feel. Whatever a new initiative, project, merger, I need to hear from someone with a passion: "This is why we are doing this!"
Rather than Nikkei reporting that Japan, belatedly, adopting the fabless business model. That's so-o-o yesterday.
This seems to me to be one more instance of a general issue in computing that has been present for decades. As technology advances, the costs of doing the required R&D, building factories, and making your own components become so high that very few can afford them, and some that can are actually joint ventures and consortiums. Increasingly design shifts to using off the shelf components and manufacturing continues to consolidate to get economies of scale.
The only was I can see this working, given the variety of technologies, design flows, and product types within the organizations to be combined is a ruthless and clear-headed assessment of where opportunities are, what opportunities make economic sense for the converged entity to pursue, and most important, what the new entity should simply *stop* doing.
Given the consensus oriented style of decision making in Japanese industry, I foresee problems. Japanese firms can execute blindingly fast once they have achieved consensus on what should be done, because part of the process of building consensus means that all concerned will agree on what needs to be done and understand what their part in the new direction is and what they must do to achieve it.
But building that consensus takes time, and will be made more complex by the fact that formerly seperate entities will be thrown together into one. Given the speed at which the industry is changing, is there *time* to come to a consensus before it becomes too late?
I think what is really needed is a leader who can simply say "This is what we're going to do.", without the lengthy effort of getting everyone to agree on the needed changes. Agreement can come later. A new direction and overall plan for getting there has to come first, and quickly.
Cynically: Maybe that is exactly why GF is suggested as part of the mix - delegate the tough decisions to a foreign entity so when the tough actions kick in, such as older fab closures and job losses it can be blamed on the culturally ignorant outsiders.
What are the engineering and design challenges in creating successful IoT devices? These devices are usually small, resource-constrained electronics designed to sense, collect, send, and/or interpret data. Some of the devices need to be smart enough to act upon data in real time, 24/7. Specifically the guests will discuss sensors, security, and lessons from IoT deployments.