A proposal to merge the system chip businesses of Renesas Electronics Corp., Fujitsu Ltd. and Panasonic Corp.—first reported by Japan's Nikkei news service Tuesday (Feb. 7)—would further the consolidation of the Japanese semiconductor industry that has played out over the past dozen years and re-shape the landscape of the country's chip industry.
According to data provided by market research firm IC Insights Inc., Renesas, Fujitsu and Panasonic had combined chip sales of about $18.9 billion in 2011, or about 6.3 percent of all global semiconductor revenue, which would have ranked the combined entity No. 3 in the world in chip sales behind Intel Corp. and Samsung Electronics Co. Ltd.
But such an analysis is oversimplified. According to the Nikkei report, the merger would only involve system chips, leaving, for example, Renesas' large microcontroller business in the company's control. Through the first three quarter of Renesas' current fiscal year, which closes at the end of March, microcontroller sales represented about 42 percent of the company's overall chip sales.
System-on-chip (SoCs) devices represented about 26 percent of Renesas' overall semiconductor sales through the first three quarters of the company's fiscal year. According to Renesas' quarterly results presentations, SoC sales represented about 27 percent of the company's total semiconductor sales in calendar year 2011, or roughly $3.03 billion, compared with roughly $11.2 billion in chip sales overall.
Fujitsu and Panasonic don't break out their semiconductor sales to the same level of detail that Renesas does, but both also have products in addition to SoCs. Fujitsu, for example, also has a strong microcontroller business, as well as memory chips, ASSPs and other devices that presumably would not be part of the proposed JV. Panasonic also offers chips like power supplies including switching and DC-DC converters, also which presumably would not be in the mix.
Removing the roughly $8 billion Renesas generated in calendar 2011 from microcontrollers and analog and power ICs leaves the combined JV with, at most, slightly less than $11 billion in 2011 revenue, which would have ranked the firm No. 6 among chip firms in sales last year, according to IC Insights data, ironically displacing Renesas.
Renesas recently advocated TSV technology for DRAM with mobile SOC's. If the rumor is correct, it's going to get spun off as well. Yet another example the adoption of new technology like TSV is suffering from consolidation.
ST-E is not looking good at the moment especially considering the mega-merge happened even earlier than the formation of Renesas Electronics in Apr2010. With the engineering teams, presumably coming with the relevant IPs, from EMP, Nokia, Philips, STM, shouldn't ST-E have the most IPR, even more than Qualcomm? If so, this is a huge competitive edge. So far, they are not winning the AP/Integration race in leaps and bounds ...
I'm a bit wary of three-way mergers; just look at ST-Ericsson! (OK they have other problems in addition) and GlobalFoundries doesn't look like a massive success at this point either. Any more word on Elpida's possible involvement?
I agree the consolidation is not just a rumor but will happen very probably.
An important aspect of this move is that Japanese auto makers including Toyota and Nissan support or possibly request it. Using the leading-edge electronics technologies is critical for the success of EVs/HVs and the 3 chip makers have been providing custom LSIs as key parts.
As long as the combined entity, if happened, does not prune the products which have very little chance to be #2 or #3, together with the related human resources, the merge can only prolong when the eventual demise is going to happen.
The involvement of Globalfoundries and the proposed manufacturing JV is the most interesting aspect of this deal to me. Assuming that the Nikkei report is accurate, I would be very interested to see how they will structure such a deal and how the manufacturing JV will be able to turn a profit.
Intellectually and strategically, it makes sense for Japan to combine Renessas, Panasonic and Fujitsu chip manufacturing into one entity. After all, it is difficult for individual companies (except Intel) to keep up with fab innovations due to costs. Organizationally, it seems it will be difficult...perhaps that is why they brought in Global Foundries into the deal. At best, it will take time to merge these manufacturing units effectively. I think this will make Japan IC chips more of a captive market within its own systems divisions. Steve Szirom, InsideChips.com.
I agree that Toshiba is generally doing better than other Japanese semis mainly due to the volume of NAND flash coming out of it's massive Yokkaichi complex, which has been built out in partnership with SanDisk. This economy of scale helps support the non memory parts of Toshiba's semi operations.
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.