AMD or Intel as well as the memory brand - page 44
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
but the Celeron 900 and e.g. Core 2 Duo E6550 are the new Intel stones judging by their colour?)
If you compare processors not by very questionable Mhz*Time parameter, but by more real Price*Time, you can get the following picture (for processors which are on sale):
Prices are taken here.
Well, not that new, but not old either (I was guided by grouping by results - that's what predetermined my colours; it turned out that all Intel's stones for people, released after Intel released multicore, have about the same ff efficiency. Strange as it may seem to you, it's the generation that makes this parameter stand out quite well). It's interesting to look at the full Q9xxx, which have 12 meg caches. They may have something close to Core ix.
Thanks for the price comparison, very curious results.
And still, Time in Price*Time parameter must be somehow unified, because of overclocking. It's better to be guided by nominal frequency figures rather than overclocking ones.
Mathemat, architecturally the Phenom II and the 6-core Opteron are one and the same. The tests on ixbt.com are largely focused on multithreading, so it seems to be lacking something, more specifically the fourth core. For our purposes Opteron or the second Phenom (even if 3-core) is the same thing.
And the full Q9xxx will perform no better than E8xxx since there those 4 cores are actually 2 E8xxx in one package and each pair of cores can use only "their" 6MB cache.
By the way I agree that you should take the PricexTime at nominal frequency, or consider the cost of the whole system (at least CPU+MB+RAM+Cooler)
Strange then why the Phenom II significantly underperformed the Opteron in both local tests. One architecture - one ff efficiency. Either there is still some serious difference, or we are missing some important factor.
At face value:
In multi-threaded applications, at this rating, AMD processors will significantly outperform Intel stones. Thanks to the extra cores.
AWESOME)
Belford, what a job) You weren't too lazy to slow down, were you?...
Personally, I'm on your side with all my hands. Time*frequency comparison turns out to be quite wrong, you squeeze an extra second less out of your CPU by increasing Hertz while a slower stone is a couple of seconds slower but has 500 Hz less math-wise wins hands down.
I think price*time along with absolute value is the most useful parameter. I would like to make two tables - for those who need maximal result of Xeon and for those who need the most effective in terms of return on investment Phenom
OK, I agree: in terms of theoretical efficiency of single-core (ff) architecture, Xeon wins (with available statistics; we haven't looked at everything yet).
But in terms of the practical "cost per unit computing speed, Price/Speed ~ Price*Time" criterion, the Phenom II X3 720 wins so far.
Probably, Athlon X2 64 will be even better, but their time has already passed: we have to take into account the absolute speed.
By the way, and new Celeron with 1 MB cache may be the first too :)
In principle, I can enter this data into the table as well.
In last place is an optetron with a xeon. My old one turns out not to be so bad, especially when working with an optimizer. :)
Yeah, thanks, joo. If the criterion calculation could also somehow take into account the time at nominal frequency (proportional to it), there would be a slight change in the totals. But that would be an estimate.