Here's one last analogy then I'll just wait to see if I get to buy Red a bottle and a steak. The first automobiles hit the road before the turn of the 20th century. In that time, we've cured diseases, made human flight commonplace, gone to the moon, invented computers the size of a breath mint, and mapped the human genome. Yet a car is basically still just a car. It goes marginally faster. It is significantly safer. It carries a few more people and is far more reliable. It has GPS and satellite radio to ease your boredom. But it's effectively the same car that it was in 1960 and 1940 and, in many ways, 1920. A to B at nominally 50-60 miles per hour. Why? Because advances in individual technologies don't automatically equate to revolutionary changes in their collective employment. The revolutionary part is too chaotic, "seemingly" random, and relies on millions of OUTSIDE INFLUENCES to make the possible gains justify the investments required. I'll just leave it at that. This book, if you haven't read it, is AWESOME and gives the background to what is driving my conviction on this (and other technology) topic. Always fun to joust with El Rojo.
I've heard about Kuhn's work for decades but never read the entire book. Conceptual philosophy just doesn't stir me like it does some folks. But it's only 175 pages. Perhaps I should put it on the list . . .
In this instance, I'm getting a 20% clock-for-clock improvement across the board (vs core 2 Quad, not to mention core 2 duo) and paying at least 200 less for a whole system. The researcher and academic in you should realize that when it comes time to get real work done, architecture trumps prettiness.
I've worked with everything from old mainframes, Unix workstations, every flavor of PC and Mac over 30 years and one thing I know for sure. Raw CPU statistics don't mean squat in getting real work done. The people do the real work, not the tools. Efficiency trumps all. Reliability, no-nonsense functionality, steady upgrades, and overall capability in a computer is what is needed to get real work done. To a computer science guy, perhaps megaflops are everything, I understand. To a researcher (a specialist in some other field) valuable time spent cobbling together computer components and software, dealing with hacking issues, and doing system administrative routines is wasted. We get paid to do field investigations and lab analysis. Creative work towards producing contract deliverables are billable hours. Time spent dinking with our machines is losing money. When it comes to saving valuable and expensive research time for me and my staff, two hundred bucks is nothing, amigo. Two thousand bucks is nothing. The loaded Mac Pro's we use cost $16,000 and are worth every penny.
