After my last article, I went absolutely off the rails and accused Micron's (NASDAQ:MU) management of purposefully running the company into the ground. Subsequent to the company's most recent chalcogenide disclosure, I no longer believe that to be true. In philosophy, a razor is a tool that allows one to "shave off" unlikely explanations. Among those, Occam's Razor often provides me with the most utility.
The explanation requiring the fewest assumptions is most likely to be correct. - Occam's Razor
Intel's (NASDAQ:INTC) management is indicating that 3D XPoint ("3XP") is the best thing since canned beer and it'll be using it in 2016 across a broad portfolio of products "from the datacenter down to low-power ultrabooks". But Micron's management indicates just the opposite; that 3XP revenue will not be significant until 2018.
There is a simple explanation for this that nobody has suggested to this point - Micron and Intel are going through a messy break-up. I've stared at the data long enough for this to make sense, so allow me to walk you through this.
3D XPoint Is Transistor-Less
I was surprised that there were no questions about fabrication size during the initial 3XP announcement. So I specifically remember when this was discussed at the Pacific Crest event shortly thereafter. While this recording has expired, the question of fabrication size was asked directly, and the answer was curious to me. It was explained that the 3XP chips have no transistors, so the traditional basis for measuring the fabrication size was impossible. But it did go on to explain that, if there were transistors, they'd be 20 nanometers in size.
This means one of two things:
- Unlike traditional memory like DRAM and Flash, 3XP chips have no integrated control logic circuitry.
- 3XP chips use some other transistor-less technology for their logic circuitry.
While I outlined in The Shining Briefcase, it is possible to use the Ovonic Quantum Control in order to replace transistors at the logic level; that is a huge leap. And there's nothing in the patent mill to show any work to this end. I think the simple explanation is that the 3XP chips don't have any logic on-board. This leaves only one conclusion.
Back when the Hybrid Memory Cube ("HMC") was initially proposed, 3D stacking was a requirement. But this requirement disappeared with the advent of much more affordable "2.5D" interposer technology. What you're looking at here (take special note of the integrated FPGA controller) is a planar "2.5D" version of the HMC. It should be noted that, back when the HMC was first disclosed, "DRAM replacement" was a goal. From the same presentation:
The propeller heads in the world of high-performance computing ("HPC") live in an academic dream world and often forget that they've signed non-disclosure agreements. They've been discussing this openly for many years now - just use a small amount of DRAM in order to mitigate the downsides of PCM.
This appears to be Intel's upcoming Knights Landing HPC processor. Also relevant from this presentation is a slide that nicely illustrates that DRAM, like Flash, is at end-of-life beyond 16Gb densities:
What Does It All Mean?
It is important to note that Intel owns the fundamental patent on the Hybrid Memory Cube (which specifically cites PCM) even though it doesn't participate in the HMC consortium. The goal here is obvious: an expensive, proprietary, Intel-only memory standard. INTC really has no competition outside of the mobile space (and 3XP will fix that), so it will necessarily command some very steep prices for its integrated products. I'd certainly line up to give it my money at this point. And everyone will, once they understand how revolutionary it is.
Intel has a lock on pretty much everything with respect to 3XP; without the FPGA control technology, the current chips are worthless to Micron. Micron will need to develop or purchase some control technology so that it can sell the chips to non-Intel customers. This is not in INTC's interest, and it is going to cost Micron a fortune, so I had expected to see an exclusive Intel/Micron joint venture announcement. Instead, we see Intel going it alone with 3D NAND and 3XP at its Dalian fab.
The Break Up
In the past, it'd just say "Intel" without beating around the bush. But INTC is now making its own chips outside of IMFT, so we know that something else is going on here.
The Rebound Relationship
When I was researching Purple Swan last June, I came across a Toshiba (OTCPK:TOSYY) job posting that discussed "Toshiba's upcoming non-volatile memory (NVM) replacement for DRAM". This was obviously curious, knowing Toshiba's affinity for phase change memory ("PCM"). But I couldn't think of anything to link this in - until now. From Micron's US patent application 20140361239:
This is a stacked PCM that is implemented using a Toshiba technology called Bit-Cost Scalable ("BiCS"). This is the technology that the industry is using to produce 3D NAND, and Toshiba generates a tidy sum from the royalties. The big advantage is that all of the memory cells in the 3D stack are fabricated simultaneously using the same fab deposition steps. And only a single bit line is required for each bit-row of stacked cells.
3XP is a traditional lithography-intense memory technology: each layer needs to be fabricated individually so it is a lot more expensive than the new 3D NAND. In a four-layer stack of cells, you'd need to spend nearly four times the amount of time to create 3XP versus this BiCS implementation. In 2012, Samsung (OTC:SSNLF) outlined that 3D XPoint (yes, it used this term in 2012) wasn't ideal from a cost perspective:
Click to enlargeIf I'm spending $5 billion on a fab and depreciating half of that over three years, this works out to more than $26/second in overhead. Certainly, there's a lot of things happening in parallel, but you get the picture: equipment depreciation is the dominant cost of a semiconductor.
When I predicted 3XP in Purple Swan, I had anticipated that this BiCS structure would be the flavor announced by Micron and Intel. When the more primitive structure was presented instead, I assumed that there may be some manufacturing or performance issue with the BiCS-style cell. I realize now that BiCS is a Toshiba technology, and there is a lot of power and politics behind it. If you investigate Micron's 3D NAND patent applications, you'll find that it is similar in complexity to the PCM pictured above.
Click to enlargeThere's no reason to produce 3D NAND or 3D XPoint if you have the rights to produce BiCS PCM. I don't believe that Intel has those rights and, as such, has only a short-term solution with 3D XPoint. It will prove very costly, but I believe that Micron and Toshiba will prevail in the long term with their BiCS PCM, which can be produced in much greater density (think 64 layers instead of two or four). Most recently, Toshiba appears to be prepping for this change. In the distance, Samsung is building a $23 billion fab just as the transistor, DRAM and NAND are hitting their end of life. I have nothing here, but I find this curious and see a potential partnership.
Don't get me wrong: Intel is going to do just fine with its walled garden packaged processor/memory/storage products. And it has many aspects of transactional memory locked up. This alone cost it a fortune - it's years ahead. And the company is doing a great job of playing a high-stakes game of chicken against Micron (see the MU share price, for reference).
I used to believe that Micron's FPGA technology acquisitions of Pico and Convey were to complement Intel's Altera acquisition (i.e. - Micron was working with Intel). I realize now that these are likely adversarial: Micron needs to build its own FPGA hybrid memory controllers in order to compete with Intel to this end. Perhaps, we'll see a Micron-Xilinx (NASDAQ:XLNX) partnership soon? Does Intel have its implementation of the HMC locked up with patents?
As outlined in The Shining Briefcase, this makes ECD's Ovonic Cognitive Computer ("OCC") entity even more key, as Micron's Ovonyx subsidiary only possesses rights to discrete, standalone PCM products. The value in transistor-less PCM is that you can integrate it into a CPU and sell billions of $10 IoT processor/memory/storage devices every year.
As a side note, I should inform everyone of my relationship with the ECD bankruptcy. I'm being kept at arm's length in order to keep me independent as a journalist. I get lots of questions about what's going on with respect to ECD, but I have no idea outside of what is posted here. Noteworthy of late is that the closing was again delayed (to March 30th). My assumption is that the sale of OCC is being wrapped up and lots of things will transpire quickly once that is finished.
My gut feeling is that Micron may be acquired by any one of a number of suitors with a contingency on the OCC deal. But I would like to ask Micron's management to remain independent as long as possible - indefinitely, if possible. This will help shareholders to realize optimal value.
I will say that I see lots of ECD technology that is awaiting the closure of its bankruptcy. Phase change window film, e-paper and display technology, for example, were recently announced by Oxford and Exeter spin-off Bodle Technology.
ECD's chalcogenide display technology - now owned by Micron - will obsolesce current LCD, LED, OLED and ePaper displays - a market of $80 billion annually. A current 4K television has just 8.3 megapixels due to the tremendous limitations of thin-film transistors. But ECD's Ovonic Quantum Control device solves all of those limitations, albeit creating a definite bandwidth issue, building displays with billions of pixels will become a possibility. Or smartphone displays that transform from backlit to color ePaper when you're out in the sunlight. The new market here is to put a couple of 4K displays on a pair of sunglasses or in an Oculus Rift. Today's TFT pixels are too big for this use case.
And it will create a new market for energy-efficient windows. Imagine a window film that can change from translucent to reflective in order to reflect sunlight on a hot day but allow it in as solar heat on a cold day. Bodle Technologies is being very secretive about who it is working with, but this Samsung PCM display patent application should shed some light:
Look familiar? This is nearly identical to the image from Bodle's website. Because Micron and Intel are bringing chalcogenides into mass production, that will place them as leading contenders to commercialize this work as well. The fundamental patents in Ovonic display are expired so, like PCM, only specific implementations will be patentable.
And then there is Ovonic optical routing. With today's technology, it takes about 75 milliseconds to get data from Los Angeles to New York. The vast majority of the 3,000-mile trip is performed over fiber optic cable. The speed of light is 186,000 miles per second so, in an all-optical situation, this trip should require less than 2 milliseconds.
There's more than 70 milliseconds of overhead in optical-electrical-optical conversion. That is, in order to traverse thousands of miles, the electrical data needs to be converted to optical. But, at each "hop" (i.e. - every major city), it needs to be converted back into electrical so that it can be routed by an electricity-based processor. With Ovonic optical routing, this does not need to happen. The world is about to get a lot smaller from a communication perspective. And Micron now owns the fundamental patents on this technology.
Compute, artificial intelligence, storage, display, communications and energy - the advent of mass-produced optoelectric chalcogenide has sweeping implications. And Micron's chalcogenide announcement reveals that this is finally happening after 50 years of research.
I believe that this will be more transformative than anything else that I will see in my lifetime. Unfortunately, a substantial portion of my net worth just expired with the January Micron calls. I realize now that this is not something that is easily understood - but is now very much an opportunity for those who can. I'm still buying Micron LEAPS with the hope that I can recover some portion of what I have squandered in ECD and now Micron.
I apologize for getting anyone involved. I absolutely believe everything that I'm putting out there. Unfortunately, I realize now that Wall Street can't possibly understand this technology until it happens.
Disclosure: I am/we are long MU, INTC, ENERQ.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Editor's Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.