SAN FRANCISCO—A potential merger between Micron Technology Inc. and Japan's Elpida Memory Inc. would dramatically redraw the DRAM landscape, creating a new No. 2 player that could challenge South Korea's Samsung Electronics Co. Ltd.'s leadership in the space, according to market research firm IHS iSuppli.
Speculation has been swirling in recent weeks that Micron is in negotiations to invest in or acquire Elpida, which has been plagued by five consecutive quarterly losses and is bogged down with debt. Neither company has publicly acknowledged any such discussions.
According to IHS, a merged Micron and Elpida would have capacity for 374,000 DRAM wafer starts per month. The combined company would have a 28 percent share of worldwide DRAM manufacturing capacity, placing it just behind Samsung, which has capacity for 433,000 wafer starts per month, or 33 percent of global DRAM capacity.
On their own, Elpida and Micron typically rank third and fourth in DRAM capacity, respectively, IHS said. The reconfigured DRAM terrain also would mean that South Korea's Hynix Semiconductor Inc. would fall to third place with 23 percent share of DRAM capacity, or about 300,000 DRAM starts per month, IHS said.
Mike Howard, senior principal analyst for DRAM and memory at IHS, said that a combined Micron-Elpida would provide Samsung with its most powerful rival yet in DRAM. "Samsung is usually 10 percentage points ahead of its next competitor, but the merger would trim its formidable market share lead by half, to 5 percentage points," Howard said.
Howard said any proposed merger or other tie-up arrangement between Micron and Elpida would face several challenges, not the least of which is Elpida's debt load. At the end of the third quarter last year, Elpida owed $4 billion in outstanding obligations, he said. While the DRAM industry is familiar with large amounts of debt because fabs can be extremely expensive to build and often require debt to finance, Micron is decidedly debt averse, he added.
The strong Japanese yen and Micron's ongoing partnership with Taiwan's Nanya Technology Corp. could also be obstacles to a possible Micron-Elpida deal, Howard said. The death of former Micron Chairman and CEO Steve Appleton Feb. 3 could also complicate matters, since Appleton was a known advocate for consolidation and was likely a driver behind any possible deal, IHS said.
But a merger between the two companies would also offer several benefits in addition to combined manufacturing capacity, IHS said. For Elpida, a deal would provide access to the premium customer segments that Micron serves. IHS noted that in the third quarter of 2011 Elpida's average selling price for DRAM was 70 cents per gigabyte, compared with $1.34 per gigabyte for Micron, a differential that IHS attributed to the firms serving different customer groups.
A deal would also give Micron greater access to Elpida's mobile DRAM buyers, IHS said. Elpida shipped 18.4 percent of mobile bits in the third quarter, compared to only 5.3 percent for Micron, the firm said, adding that mobile DRAM now accounts for about 15 percent of the total DRAM market.
For customers of both Micron and Elpida, a merger between the firms could be a double-edged sword, IHS said. A merger would mean one less DRAM supplier on the market, which could enable DRAM prices to firm, IHS said. But a merger could also provide ballast against the overwhelming size and influence of Samsung, IHS said.
Overall, the benefits of a more stable market resulting from the merger probably would outweigh the drawbacks of having one less supplier, Howard said.
I'm no financial whiz, but there has been plenty of examples of companies buying up other companies that are bleeding money. This almost always does not end well. It seems like the time to buy is after they've already bled out (i.e., in liquidation mode). This would also raise the spector of more US jobs lost.
Yeah, this deal doesn't make sense to me either. Just let them die, then feed off of their carcass. Viola! One less competitor for free. There is definitely not $4B+ worth of value there, and that's way too much to pay just to say you're losing less than everyone else in DRAM.
I'm not sure if this merger really can bring advantage to Micron at the end. It seems the only likely scenario is that the profit margin of Micron will drop further while it is still hard to say if the price can stand firm.
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.