|
Random photons, scattered electrons, and cathartic rantings from SPIE and beyond |
|
|
|
Mar 06, 2008 at 08:30 AM |
When Micron announces fab plans, don't hold your breath.
So Micron has said it will build its next fab in the Boise area, although the company hasn't announced any sort of schedule. Does anyone remember how long the company's fab shell in Lehi, Utah, stood empty and unused? Pretty much a decade, give or take, a classic case of adding capacity when it wasn't needed (or at least a building which could house added capacity). The shell was the butt of jokes (and focus of speculation) for years, referred to by more than a few souls as "the Barn" or "the Hangar." Finally, last year, the Lehi facility properly ramped to volume production, as part of the IM Flash NAND venture jointly operated by Intel and Micron.
This time, I hope Micron knows better than to build it, wishing they will come, when there's no one on the way.
Yet another three-letter acronym. One of my favorite initialisms from this year's SPIE Advanced Lithography comes courtesy of Cymer, the laser light-source folks. It's ABC, which stands for "advanced bandwidth control," which sounds like something we could all use from time to time. The "added knob," as the company's Nigel Farrar calls it, is part of Cymer's high-performance upgrades to its XLA systems. Minding one's ABC helps keep the process window tight, even with long-term drift, in the chamber, chamber to chamber, and tool to tool. The operator can match the bandwidth from source to source, which will help keep the process dialed in and smooth out the move to high-volume immersion litho.
Now if only I could get my own bandwidth issues under control, and stop drifting when I should be focusing, but being creative I do like to drift...now where was I?
More adventures in terminology. Kurt Ronse, IMEC's litho pointman, used the term "triangle of death" (or "desperation," if that's your druthers) to describe the advanced photoresist trifecta of sensitivity, resolution, and line-edge roughness (LER) that "is a generic problem for all the optical wavelengths." If you crank up the sensitivity, you take a hit in resolution and can't push past 30 nm. To deal with LER, you can use acid diffusion to smooth out those rough edges, but that degrades both contrast and resolution. "There needs to be some bright ideas" to find a solution, he said, adding that there are synergistic benefits of trying to defeat the deadly resist triangle, since "it will benefit both advanced double patterning and EUV. "
Sign of a fading company? In what could be seen as poignantly symbolic, Canon's SPIE booth was a simple 10 x 10, a far cry from the pavilion presence it had at the show during its salad days as a market leader. Although the company did present papers on its immersion and EUV platforms, several sources told me a customer that received one of the Canon IML tools shipped in November has decided to return the system to the equipment supplier. If true, the rejection reminds me of the old saying about adding insult to injury.
New gig. When I first ran into Ken Rygler at SPIE last week, the seasoned litho and mask veteran handed me a new business card. Ken is the freshly minted chief marketing officer, or CMO, of Molecular Imprints. He told me it's a contract gig for the time being, so his Rygler & Associates consultancy remains in business. I reckon we'll see mo' of Ken on da nanoimprint litho front.
Speaking of nanoimprint...or not. In his SPIE report for Solid State Technology's Website, industry consultant Griff Resor describes what he sees as the litho options that the industry faces at the 22 nm half-pitch node, mentioning enhanced optical litho (double patterning, high-index immersion, and such) and EUVL (he thinks the big guys will use it for the trickiest mask layers) as the likely candidates, with e-beam direct write potentially in the running. I guess the Griffster missed or chose to ignore the numerous papers and announcements that point to another, increasingly viable 22-nm litho contender--nanoimprint---since he didn't even mention NIL in his article!
Founder of Asyst tries to buy back...Asyst. Many of us attending SPIE remarked on the sheer weirdness of last week's news about an investor group led by the Gores Group and Aquest Systems, the new kid on the automated material handling systems (AMHS)/fab automation scene, launching a hostile takeover bid for Asyst (an effort so far rebuffed by Asyst's board). Why the head-shaking double-takes? Because Mihir Parikh, head honcho at Aquest, was the founder and CEO of Asyst for years and left the company back in 2002, and Menlo Ventures, one of Aquest's private investors, was Asyst's founding investor back in the day.
I don't get it. Haven't Mihir and Menlo been there, done that? Does he miss his "baby" that much or think that Asyst is so poorly run that he must try and regain control of it? How long has he been thinking/scheming about this takeover ploy? I haven't spoken with Mihir about his motivations, but one word commonly used to describe many Silicon Valley execs comes to mind---ego.
|
|
