XBOX 360 2714

snip

Major thread drift here -- but why not, since none of what's being discussed has more than a very slight relationship to the subject line anyway .... ( :-)? ):

The book (*) my school uses to teach our combined course on buttembly language, computer architecture, and (a bit of) circuit design has a nice chapter on performance that makes exactly this point -- that clock speed does not tell the whole story. (I would describe the something-Hertz number in terms of, oh, "lowest-level operations per second" rather than in terms of electron transport, but maybe I don't understand what you mean by that.)

((*) Patterson and Hennessy's "Computer Organization and Design: The Hardware-Software Interface". I haven't checked the most recent edition, but previous ones had this chapter, and surely it's been kept.)

The discussion (in the book) focuses on what factors affect how long it takes to execute a particular program, and those factors include clock speed, but also things such as cycles per instruction, instructions per line of high-level code, etc. It's an interesting discussion, and I *think* students do come away realizing that the impressive something-Hertz number quoted for the chip doesn't tell the whole story.

But that's still a bit low-level for what you're asking, maybe. I'm not sure how one could compare different systems with regard to user productivity, but it would be an interesting and valuable thing to do, if one could .... ?

-- B. L. Mbuttingill ObDisclaimer: I don't speak for my employers; they return the favor.