2014: That was a year that was! Many of you, perhaps remembering my outlook missive from last year, must be looking forward to how I would explain my piffle given the marvelous developments over the past year. Well, piffle it ain’t. Just more nuanced. I’ll explain but first, a recap of events:
2014 clearly belongs to Samsung.
- First, in February, they came with their ISSCC presentation with a 24-layer vertical channel 3D NAND chip.
- Second, around midyear, there was the rejoicing around their 850 Pro SSD incarnation of their 32-layer product chip, which generated a lot of technical and market analyses.
- Third, in November, Intel trumpeted their 32-layer 3D NAND technology.
- And fourth, akin to Sherlock Holmes and the “curious incident of the dog in the night-time”, we didn’t hear much from Toshiba/SanDisk and Micron. Indeed, Micron came out in the New Year with an almost hand-wringing account of the same technology as Intel, the significance of which was discussed in depth.
I was clearly wrong to say that these “litho-light” 3-D Flash approaches would not appear. They certainly did and in product form too.
However, the enthusiasm for this approach (all these vertical channel approaches are very similar but no doubt each manufacturer will make use of the word “revolutionary” for any slight differences that may arise) has evolved into three forms: the “foaming-at-the-mount” excitement of Samsung and Intel; the tortured, caveat-full statements from Micron; and the silence from Toshiba/SanDisk. Why all this confusion and mixed messaging from those in the real know?
The answer is cost. Or at least the road to low cost. The arguments for this have not changed since my 2014 outlook. To get closer to being cost effective, the manufacturers have three paths open to them: multiple electrical bits per cell; longer (more stacked layers) strings; and lateral shrinking of the cells. Clearly, the first is the easiest but the other two remain and are rocky roads indeed.
What is my position now? Following George Orwell’s advice on this one (in a nutshell, use fresh images and simple language), many of you will have marveled at the recent free climbing ascent of El Capitan in Yosemite National Park. It’s the same awe we have for those engineers who have built Samsung’s V-NAND product. This awe is independent of the knowledge that there are simpler walking paths to the summit with less cost and injury to fingernails and muscles. We at Schiltron believe we have one such path.
Another apt analogy is the peacock’s tail. What a marvel of complexity and beauty! But when compared to the swallow with its ideal form adapted for fast, low cost flight, perhaps that tail needs to be adjusted a bit.
My predictions for 2015 are the following:
- Samsung will come out with a close-to-50-tall structure.
- Intel’s foaming will lead to the almost inevitable SSD appearance of their chip.
- Micron will dip more than a toe into the water by giving more technical details of their structure (including the so-called “revolutionarily” different ones from Samsung’s approach but the same thing as Intel’s).
- Toshiba/SanDisk will push 2D further by shaving off more nanometers to get to “sub-15nm” but still be reticent on their 3D approach.
And, analogous to a famous advert for “beer that refreshes parts that other beers cannot reach”, perhaps, just perhaps, another approach may break through that reaches lower cost that others cannot reach.
Let’s compare notes in a year’s time. ~ AJW
