Over the last couple of years, chipmaker Intel (INTC 1.74%) has certainly become a more aggressive player in the world of non-volatile memory technologies. The company has long built and sold NAND-flash-based products targeted at some parts of the personal-computer market, as well as at the data center. The company has worked with memory specialist Micron (MU -0.94%) to develop memory technology, and had long held minority stakes in NAND-flash manufacturing plants co-owned by Intel and Micron.
Image source: Micron.
In late 2015, though, Intel announced that it would be converting its logic-chip manufacturing plant in Dalian, China into a NAND-flash and 3D XPoint memory manufacturing plant owned entirely by Intel. Further, Intel has indicated to investors on several occasions that it has been stepping up its research-and-development investments pertaining to next-generation non-volatile memory technologies -- both NAND flash, as well as future 3D XPoint iterations.
Given the growing demand for NAND flash as the world transitions away from mechanical hard-disk drives, as well as the opportunity for Intel to attack the traditional data-center DRAM market with 3D XPoint-based memory modules, these investments seem prudent and could pay off handsomely in the future. However, despite Intel's newfound interest in non-volatile memory, the company isn't being aggressive enough. Allow me to explain.
Content share-gain strategy is obvious
Most of Intel's NAND flash-related revenue comes from sales of NAND flash-based storage drives to major data-center customers. These drives combine lots of expensive NAND flash with Intel's controller technology -- both hardware and firmware -- but data-center operators are willing to pay handsomely for these products, as they can often deliver substantial reductions in data-center total cost of ownership compared to mechanical hard-disk drives.
The company also sells NAND flash-based storage drives into the consumer market, but the company isn't really a big player here. The company's storage drives sold individually to consumers tend to be high quality, but they're often not competitive on price per gigabyte or even price to performance. Intel NAND flash-based storage drives are also a rarity in pre-built personal computers.
Obviously, on a per-device basis, consumer-grade NAND flash drives aren't anywhere near as lucrative as high-end data-center-oriented drives that sell for hundreds, if not thousands, of dollars. However, the number of personal computers that are sold each year is in the hundreds of millions, and most of those computers come with Intel processors.
One way that Intel has been trying to combat the secular decline of the personal-computer market has been to try to boost the amount of dollar content that it sells into each personal computer. Intel has done this in a couple of ways, from simply bundling auxiliary chips with its platforms -- e.g. Ethernet and Wi-Fi/Bluetooth chips -- to outright integration of functionality into its processors -- e.g. sensor hub, graphics, memory controllers, and much more.
If Intel were willing to make significant incremental bets in NAND flash capacity, and if it were willing to price aggressively enough, it could offer NAND flash to its personal-computer customers/partners as part of a broader bundle, including other chips. The profitability per unit of NAND flash might not be particularly high in this case, but Intel would certainly be able to leverage its position as the key vendor of personal computer platforms to sell a lot of NAND flash at a reasonable amount of profit per unit.
That could add up to a sizable boost in revenue for Intel, and it could create a positive feedback loop vis-a-vis the company's memory strategy. The more NAND flash revenue the company generates, the more it'll be able to reinvest in the development of newer, better, and more cost-effective NAND flash technologies, which could, in turn, fuel further NAND flash revenue share gains.
