| |
AMD working on new PC rating method. CNET Aug 21 2002 6:03PM ET [Moreover - moreover...]
The problem with creating a new figure of merit is that no one will trust that you aren't just trying to make yourself look better. And of course if you didn't need to look better you wouldn't be interested in a new FoM in the first place. Transmeta found this out when fishing for a metric that accounted for power as well.
The reason for making something different, in any field of endeavour, is that you see a legitmate need for something that traditional methods aren't covering. But when you cry "Spin" too frequently that's all people come to expect from you. The race to 1 GHz was where this comes from, and it was fine with AMD... as long as they were ahead.
1:23:10 AM  Categories: Pushing rectangles...
|
|
Graduate Students and Tuition: Dave Fletcher comments on the Utah Legislatures fixation on out of state graduate students:...
He continues with some good comments on the marginal cost of admitting new students. My take is slightly different: Here are a bunch of very bright people who are willing to come to our state and work for almost nothing in a university research program. Basic research and number of patents are one of the very best predictors of economic prosperity for a country (and probably a state as well). We ought to let them come here for free! The fact that we're able to take back almost all the trifling amount of money we pay them as research assistants in fees and tuition is amazing. [Windley's Enterprise Computing Weblog]
People propose that graduate assistants pay full tuition. How many propose that they should be paid market wages? I've had 20 TAs consume 60+ hours a week, then you go to classes. And THEN you do some research to coble together a thesis to get out.
1:08:17 AM Categories: Pushing rectangles...
|
|
|
