According to report, Samsung has a plan to aggressively expand its foundry capacity from its current 27,000 eight inch equivalent wafers a month to 40,000 by the end of the year, to 125,000 by the end of next year and to 200,000 by the end of 2012.
This, of course, is a long way behind market leader TSMC’s 10m eight inch wafer equivalent capacity a year. Nor does Samsung appear to have a strategy to build 100,000 wafer a month GigaFabs to acquire the same economies of scale as TSMC.
Recently, Ana Hunter, Samsung’s foundry vice president, told EW: "It’s a very important business for us - a strategic growth engine – we entered it after several years of study about what could provide significant growth for Samsung."
Asked why Samsung sees foundry as a growth driver at a time when TSMC is finding growth so elusive in the foundry business that it is diversifying into photovoltaic and LED, Hunter replied:"Samsung is relatively small as a foundry and we see opportunities for growth. We think the market wants choices at advanced technology nodes. There are not many who can provide 32nm, 28nm and 22nm – we see a market for that."
Samsung’s foundry strategy is to go for the high-priced, high-end, advanced segment of the industry.
"We focus on advanced technology nodes – 45nm and prototyping on 32nm," said Hunter, "we have customers taping out on 32nm in Q3 on our high k, low power process for mobile applications. Risk production on 32nm will start in a few months." Samsung has foundry customers taping out on 32nm, said Hunter.
Samsung is looking for more foundry customers. "We’ve started with a handful of companies we want to work with including Qualcomm and Xilinx," says Hunter, "and we’re looking for additional customers who want advanced technology nodes." Another customer is Texas Instruments.
With Glofo (Globalfoundries) also building capacity, the current shortage of foundry capacity may not extend beyond the end of next year.