The industry appears to be caught in the grip of buyout fever. A week ago, DRAM was making headlines with the Micron Technology Inc.’s move to acquire Elpida Memory Inc.
Now, the focus has shifted to the non-volatile space with the news of EMC Corp.’s $430 million cash acquisition of Israeli flash memory startup XtremIO Ltd. That's a lot of money to pay for a three-year-old company that’s essentially still prerelease, but EMC is well positioned to do it, with 1Q12 income of $587 million. It’s a gamble, but if the technology succeeds, the storage manufacturer will be well positioned to sell SSD solutions into the developing cloud market. "One of the issues with flash technology is that there's a wall out there where scaling is going to stop working for NAND flash,” says Jim Feldhan, President of Semico Research Corp . “That wall has moved over time but I think now we have a general consensus that the 10-nm or 8-nm nodes will be the end. Companies are really looking for innovative ways to keep the technology going and that becomes very valuable if one of these technologies works out.”
According to XtremIO’s website, which is long on general statements but short on details, the company focuses on flash data storage solutions for ultra-high-performance enterprise server applications. Not surprisingly, company doesn’t figure among the top current NAND flash vendors, which consist of Samsung (33.9% market share), Toshiba (32.9%). Micron (14.2%), Hynix (11.5%), and Intel (7.6%), according to the latest numbers from TrendForce. In other words, the buy isn’t likely to alter the non-volatile memory market, at least not in terms of cash and bits. It may have an effect on the storage landscape, though. Indeed, if XtremIO’s technology lives up to the hype, EMC may have made a very savvy buy.
The company’s most likely strategy is to hold onto the technology for internal use rather than selling flash solutions itself. Given that total revenue for NAND flash alone dropped 2.5% quarter-over-quarter in 1Q12 to $4.79 billion, despite a 16% jump in bit shipment, keeping higher on the food chain is probably a good thing. “If they take that technology and incorporate into their end applications, there’s much more value there,” says Feldhan. “[EMC’s] can afford to pay to have a more expensive technology because they're getting a higher value for their end product versus just selling the nonvolatile memory and playing the cost-per-bit game.”
The deal is representative of a sea change underway in the memory industry, perhaps because the technology is approaching an inflection point. “For years we've seen companies in the industry divest out of technology,” Feldhan observes. “They spin the technology off and they focus on their end application. Now, the pendulum is starting to swing the other way—companies are starting to vertically go down the supply chain to the component level.”
Of course, if the XtremIO technology is all it is made out to be, it could put EMC a power position – at least for a while. The reality is that even with patent protection, they might remain vulnerable to competitors taking over the technology. It comes down to not just the performance of the devices but to the strength of the patent portfolio—weak patents can be broken or supplanted. Strong ones offer protection, at least to a point. Texas Instruments, for example, held seminal DRAM patents. That didn't stop any of its competitors from developing DRAM but they had to license the technology from TI. “At one point, it was one of TI’s biggest revenue streams,” Feldhan observes.
Did you find this article of interest? Then visit the Memory DesignLine
where we update daily with design, technology, product, and news
articles tailored to fit your world. Too busy to go every day? Sign up
for our newsletter to get the week's best items delivered to your inbox.
Just click here
and choose the "Manage Newsletters" tab.