Brian's Brain

Aug 29 2007 8:00AM | Permalink | Email this | Comments (9) |
Blog This! using:  Blogger.com | LiveJournal |
Digg This | Slashdot This | add to Del.icio.us

Before time further slips away ;-), I want to take a few minutes to respond to a comment 'slacker711' left me three months ago, on the eve of the SID conference.

I'll be curious to hear your comments from SID…particularly as to how you see the AMOLED market developing. There has been quite a bit of hype lately but, as usual, the ability to drive down costs is going to be the limiting factor...

Organic light-emitting diodes are one of those seemingly perpetual 'almost-here' technologies that I regularly encounter, a list that also includes such notables as:

• reconfigurable computing
• GPGPU, and
• MRAM

What OLED has in common with its above peers are (at least) two high-level factors:

• unbridled promoter optimism that the technology will not only create new markets but also, by virtue of its technical attributes, substantially-to-completely displace now-dominant technologies in existing markets (in OLED's case, the liquid crystal display aka LCD), and
• perpetual under-performance to that promoter optimism

Don't get me wrong. I actually buy into the 'new market creation' aspect of the OLED hype, although the technology's backers and I probably disagree on the 'when' part of the pitch. OLED's self-illumination characteristics translate into no need for a backlight, as is the case with a LCD. OLEDs don't require LCDs' intensity-sapping polarizing layers, either. Couple these two factors with OLED's unique manufacturing process, compatible with both inkjet and screen printing approaches, and you have a technology that's applicable to a wide range of materials, including flexible substrates.

But, as plasma backers are also learning, LCD is the '800 lb. gorilla' of the display industry, garnering a tremendous amount of R&D and manufacturing attention from a large number of suppliers. Only a few years ago, few observers thought that LCD would ever substantially close the gap on plasma with respect to black levels, viewing angles or other measures of image quality. Today, although advanced plasmas still hold a tenuous edge, their lead has largely evaporated. And LCDs' Moore's Law-reminiscent cost trends have rendered them price-competitive with plasma at 42" and smaller screen sizes, as my early 2005 feature article predicted, and especially when the displays' native resolutions are factored into the mix.

These and other LCD improvements have also enhanced the technology's competitiveness versus OLED. LCDs may not have sub-msec response times, but their few-msec switching speeds are arguably 'good enough' for a wide range of applications. LCDs still require backlights, but they're now much thinner, and the in-progress transition from cold-cathode fluorescent (CCFL) to LED backlights will further boost that LCD-slimming trend as well as diminish OLEDs' color gamut and power consumption advantages versus LCDs. And let's not forget about OLED's lingering lifetime issues; arguably they aren't an issue with rapidly-replaced cell phones, but generally speaking I'm not fond of any design approach (such as a non-replaceable embedded rechargeable battery) that ensures premature system mortality.

Yes, I know about iRiver's Clix, Creative Labs' ZEN MicroPhoto, ZEN Sleek Photo, and ZEN V, and the (few) other color OLED trendsetters that the display technology's backers regularly tout. And yes, I also know about the larger list of monochrome OLED adopters; various cell phones (but only for the secondary display) and Bluetooth headsets, Creative's ZEN Stone Plus, Sandisk's Sansa Clip and Sansa Express, D-Link's DIR-660 Limited Edition router, etc. But I'd argue that in these cases, OLED hasn't won by virtue of any particular feature advantage aside from, perhaps, a manufacturer's desire to differentiate a product by means of monochrome OLED's unique black-text-on-blue-background appearance versus a more staid traditional monochrome LCD.

Instead, I think that fundamental price-and-availability business factors are at work here. I'll explain by means of an analogy. Back when I joined Intel's nonvolatile memory group in 1997, the company's flash memory products spanned a 256 Kbit-to-2 Mbit density range, and Intel was happy to sign up whatever business it could even at the low capacity end of that range. A few years, a few lithography shrinks and a few wafer stretches later, and a 256 Kbit flash memory die was so small (and so numerous on a die-per-wafer basis) that no amount of customer demand could come close to filling a fab (an aspiration which many of you probably already realize is critical to a sustainable semiconductor business).

By the turn of the century, even the high volumes of the PC flash BIOS market were insufficient for the fab-filling aspirations of Intel, which had sharpened its focus on the cellular business and on multi-Mbyte per-chip opportunities. Intel therefore signed up alternate-source vendors like SST so that all-important PC customers wouldn't be left high and dry without supply. And more generally, EEPROM and other nonvolatile memory technologies have filled in the vacuum that Intel and its NOR and NAND flash memory competitors create when they exit a particular density threshold.

As I earlier mentioned, LCDs' manufacturing characteristics are very semiconductor IC-like, and therefore I suspect the same obsolescence factor is at work here. Tiny LCDs for cell phones and portable multimedia players don't fill fabs near as effectively as massive computer and television monitors (for which business seems healthy) do. So I suspect that the TFT-LCD suppliers are deliberately pricing themselves out of the few-inch display market, thereby handing that business over to OLED (along with, to some degree, trailing-edge STN-LCD).

Agree or disagree, folks? I 'spect I'll get at least a few OLED enthusiasts who take issue with my stance, but regardless of your opinion I welcome your feedback.