On July 28, Intel (INTC +1.07%) and Micron (MU 0.22%) jointly announced a next generation technology that they dub "3D XPoint." The technology, according to the press release, is a new type of non-volatile memory (i.e. the data remains in memory even when the device is no longer powered on) that delivers some major improvements over existing memory technologies.
According to the companies, this technology delivers up to 1,000 times lower latency than traditional NAND flash used in solid state drives today and is 10 times denser than conventional DRAM memory.
3D XPoint is, according to the companies, expected to sample to "select customers" later this year, and both companies say that they're developing "individual products based on the technology."
Following this announcement, shares of Micron rose 9%, handily outperforming the iShares PHLX Semiconductor ETF (SOXX +2.10%). Intel rose 2.15% for the day, performing roughly in-line with the aforementioned semiconductor sector ETF.
Let's take a closer look at what Intel and Micron announced today.
A "new class" of memory
Intel and Micron stressed that this technology spawns a "new class" of memory and won't necessarily make obsolete the other types of memory technologies. I believe DRAM will still be faster and thus a more suitable technology for system memory for most devices (although Intel/Micron did indicate that 3D XPoint could be used as system memory).
A wafer of Intel and Micron's new 3D XPoint technology. Source: Intel.
I also get the impression that traditional NAND flash will be more cost effective, making it the technology of choice for more traditional, more cost-limited storage end markets.
Where I think this technology has the potential to shine is in data center applications. In these markets, customers tend to care about total cost of ownership rather than just upfront acquisition costs. This means that if the improved performance of this 3D XPoint technology can ultimately yield a total cost of ownership benefit even if upfront acquisition costs are higher than traditional NAND solutions, then the solution could be quite successful.
Is this a game-changer for either Intel or Micron?
This technology sounds very interesting, but I'm not convinced that it will be a financial game-changer -- at least in the near-to-medium term. For Intel, I suspect that the main benefit of this technology will be that it can offer even more compelling server platform solutions to customers, helping to further widen the moat that it has around its highly lucrative data center business and capture additional revenue content.
Micron, which focuses primarily on developing, manufacturing, and selling memory, could also see a revenue boost to the extent that this technology takes off.
However, as an investor in both Intel and Micron, I would be very interested in getting a sense of how large the market for this 3D XPoint technology might be in the near-term and what kinds of growth rates the two companies expect from this technology.
Micron is expected to host an analyst conference on August 14, so perhaps investors will be able to glean some insight into what the two companies expect from this technology. And, if Micron is light on the details at that event, Intel executives might shed some additional light on the financial expectations for this technology at its own investor meeting which is expected to be held in November of this year.
It's not yet clear if this technology will ultimately prove a "game changer" for the companies (as one analyst in the audience pointed out, there have been other potentially "game changing" memory technologies announced that never saw the broad adoption some had hoped for) but from the announcements thus far, the technology seems pretty promising.
