How can you possibly say that Leinies is the official beer of supercomputing after they were acquired by Miller, the paramilitary branch of Phillip Morris. Leinies is not what it used to be - spiritually or tastewise.
If Huber won't work for you, how about Red, White & Blue or Grainbelt? Maybe Andeker? Even Old Style (aka Dog Style) will work in a pinch. Wish I could email you a sixer, but that protocol has not been released yet.
Off topic, what they hell is going on with our little company here, Third? I was hoping for some press releases after the new year to take advantage of this January Effect. Any insights or guesses?
Suitable - please pick up the white courtesy telephone and give us an update.
Being the reasonable fellow that I am, Point Bock it is.
The Official Beer of Supercomputing
Does this mean that most supercomputing projects will be scheduled in the Fall?
I note the absence of Masked and Preston, who were lipping off in December. They are very quiet now. Suitable has also gone dark. Pointdexter is missing in action. And I am getting concerned - I want to hear the SD acceptance and the European order, not more handwaving about CMOS. Deliver now with what you have, we don't want more promises of how great the future will be.
(mysteriously around the same time Tera was founded) when they sold out to those executives from Pabst, who ran the place into the ground and sold off the Augsberger line to Stroh's. While the family is back in charge now, they're hardly in a position to take on a responsibility like Official Beer of Supercomputing. In any case, I saw Burton at the Blue Star a few weeks ago, and while I didn't go over to ask what beer he was drinking (or when they'll get four stable processors) I can tell you authoritatively that the Blue Star serves neither Huber nor Leinenkugel's.
Can anyone tell me what a 'test chip' is, regarding the development of the CMOS machine, and whether having one represents significant progress or not? Is this one of the first stages of the journey or the last?
> Can anyone tell me what a 'test chip' is, regarding the development > of the CMOS machine
Test chips are built to prove basic elements of a new technology where there are significant unknowns. If test chip reveals problems in the technology, there is time to redirect the primary design effort to avoid trouble without a major schedule slip. Typically a test chip does not include a significant amount of functionality from the actual design.
> .. having one represents significant progress or not? > Is this one of the first stages of the journey or the last?
If you're talking about a Tera test chip, it certainly represents a big step up in risk management. One could only guess (and then multiply by 5) at what it means for delivery of a production machine. Jim, Do you know whether the chip is back in Tera's lab? Is it a full size die?
from http://biz.yahoo.com/bw/981021/tera_compu_1.html > Key to Tera's choice of TSMC as its foundry is TSMC's singular > ability to deliver the 3.61 square centimeters die on a 0.25 micron > process critical to CMOS technology, while meeting high yield > criteria.
Just proving that they could build a working device of that size would be noteworthy. 3.61 square centimeters would be about 3/4" on side which is pretty damn big - enough for maybe 9 or 10 million gates. That, by the way, is 100 times more gates than they are getting on the current MTA GaAs chips. It's hard to beleive, but the actual die sizes are not that much different; its the density of CMOS that has improved in the 10 years between the two technologies.