| PLEX86 | ||
|
dual processors and linux 1776Roy L. Fuchs Back when I worked at a company that designed and manufactured CPUs, the CPU speed given was the actual clock speed of the chip, so the number was accurate. I imagine the specified clock speeds of Intel Chips, when they specified them, was accurate as well. For our chips, that meant that the fastest operation (probably clearing a register or something) could be done in one tick of the clock. Now the thing is that clock speed, in and of itself, does not mean much. For example, these days, lacking a fast enough, large enough, on-chip cache, the whole machine would be limited by what is now called the "front side bus", or the memory speed, which ever is less. And for IO-limited processes (such as much database work), even that does not matter as much as the IO system and its various choke points. BTW: most of this discussion is not answering the original question of "What are the operating system issues that determine application performance with dual processors? " dual processors and linux 1777 BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I hereby accuse Michael Heiming of stating: This is indeed intentional. Back around the time when AMD introduced their XP... Dave's Tinkering Again Hola El Groupo! Well, here I am again .. you know by now after my 2-1-2 years of postings what a tinkering kind of guy... -- .~. Jean-David Beyer Registered Linux User 85642. V PGP-Key: 9A2FC99A Registered Machine 241939. ^^-^^ 07:05:01 up 84 days, 20:37, 5 users, load average: 4.45, 4.39, 4.27
|
||||