| Merry Christmas 2008! | 
|
Home
| Merry Christmas 2008! |
|
Well, I think I speak for most folks when I say "2008 was a rather interesting year." As in, interesting in the sense of the famous Chinese aphorism (which is really a curse in disguise) "May you live in interesting times." Decades of leaving beyond our means, encouraged by banks, credit-card companies, Wall Street and the "cheap credit" policies of the Greenspan-led U.S. Federal Reserve have finally caught up with us Americans (and folks in many other countries as well), and the hangover is proving to be nasty. On the positive side, it seems that many people are doing what people are wont to do in hard times, namely learning to refocus on what really matters, meaning things like friends, family, health, having a steady job even if it's not the ever-elusive "perfect job", and so forth. If the developed world emerges from this financial crisis - and make no mistake, that could take many years - in a slightly less materialistic, selfish form, that will make up for much of the present pain, which I hope is not hitting any of the readers of this missive in all-too-harsh fashion, although I fear that may be too much to expect. But it was a quite interesting year in other, more personal ways, so without further ado, on to the yearly roundup.
Well, it's been nearly a year-and-a-half since I joined Synopsys and I am pleased to report that the marriage has been a happy one. Early this year I finally got my "real" work assignation, as one of the software developers in the newly-constituted Schematic-Driven Layout (SDL) group of the Custom Designer (CD) project. CD is Synopsys' new flagship product, designed to compete head-to-head with the longstanding industry leader in the area of custom chip design, the Virtuoso tool from Cadence Design systems. We formally debuted Custom Designer in April, and early customer reviews have been extremely positive. That of course means that we now have to try to satisfy the very high expectations of our new and prospective customers, so we've been very busy. Our little group has jelled nicely, and it's a pleasure to work with colleagues as talented and dedicated as the ones I have. While Synopsys has not proved immune to the overall economic downturn, we are in far better shape than most of our competitors, and so far have had only a tiny number of job cuts, the main economic fallout being a hiring freeze and some strategic reshuffling of various business units.
Alas, in typical Silicon Valley fashion there wasn't all that much of this during the year, but Mom came out to California no less than three times - once around Easter, once more in mid-October, and then also at Thanksgiving - and my sister Ingrid and I did our best to show her a good time. For Easter we took a trip along the coast to Half Moon Bay and had dinner on the terrace of a lovely seafood restaurant overlooking the Pacific. In October we spend several days in Yosemite enjoying the Fall colors - that's a picture Ingrid took from the Valley floor during one of our hikes - and for Thanksgiving Mom and Ingrid had some shopping and tea-at-the-Ritz fun in San Francisco, and then we all got together with a bunch of Ingrid and Ian's friends in San Rafael for Thanksgiving. Ingrid and Ian are expecting - twin boys, no less, so "two for the price of one!" as the merchants like to say on their sale advertisements - in March, so I'll probably be spending a fair amount of time with them at their homestead in San Rafael in the coming years.
I would be remiss if I neglected to share the news highlights related to my favorite hobby - and in reply to the rhetorical question of "What's new on the prime-number front?" I am pleased to say that this year the GIMPS project (mersenne.org) discovered not just one but *two* new world-record-sized Mersenne prime numbers.
Reproduced above are the last few lines from the run-status file resulting from Sun's Tom Duell running my Mlucas program on 16 processors of a supercharged Sparc VII server to verify the new discovery. It may not look like much, but when you think that every one of those 43112607 iterations (which equals the base-2 exponent of the new prime, minus 2 - that's the "magic number" of steps the famous Lucas-Lehmer prime test requires) of the test of that Mersenne number involves exactly multiplying together two gigantic integers having nearly 13 million decimal digits each, you get a better sense of the amount of computational effort required. This is the first known prime number having 10 million digits or more, and as such, qualifies for a $100,000 prize from the Electronic Freedom Foundation (eff.org), to be shared by the lucky person whose PC happened to have the "lucky lottery ticket" (that is, the one prime exponent among the tens of thousands which have been tried since the last Mersenne prime), Edson Smith of UCLA's mathematics department and GIMPS founder Goerge Woltman. The new discovery also earned the #29 spot in Time magazine's Top 50 Best Inventions of 2008 - the picture at right is from the Time article.
Now one new world-record prime would have been interesting enough, but 2 days before the above double-check of 243112609-1 finished, a *second* new prime popped up on the Primenet server - this one was in fact smaller than the first prime but was also larger than 10 million digits, so the finder of that one missed a nice payday by only a week. As it turned out, the fellow whose PC found the second prime (Germany's Hans-Michael Elvenich) is an environmentally conscious sort and only ran his test during off-peak hours for electricity demand. We hope his just missing out on the money prize does not dissuade him from his otherwise-laudable green habits. And here's the official double-check confirmation of the second discovery, again run using my Mlucas code, this time on a spiffy even-faster Sparc VIII workstation at the Sun offices in Menlo Park, California:
Merry Christmas/Happy Holidays and a good year 2009 to all, and to our European friends and relatives, Frohe Weihnacht und einen guten Rutsch ins Jahr 2009!
