AMD servers use more power then their Intel equivalents.

[zekrahminator]

Intel recently tested their brand new Xeon processors for power consumption, and pitted their Xeons against the previous reigning power consumption champion, the AMD Opteron series. Intel put a Xeon 5160 based system (2x3.0GHz, 8x1GB) up against an AMD Opteron 285 processor based system (2x285, 8x1GB). The "Intel sponsored" results showed "curiously high" power consumption for the AMD Opteron system. And so, AMD decided to run this test on their own, using their own testing environment. Surprisingly, they not only got similar results, but the Xeon used less power in the AMD tests then they did in the Intel tests.
http://www.techpowerup.com/img/06-09-07/amd _power_thm.jpg (http://www.techpowerup.com/img/06-09-07/amd _power.jpg)

Source: TheINQ (http://www.theinq.com/default.aspx?article=34231)

[XooM]

Erm, how are those results similar? In Intel's, the opties use more. in AMD's, they use less.

by the way, whats an "Opertron"? :p

[wazzledoozle]

But the Opteron still has lower idle and TDP power draw than the Xeon's. :)

(Although the Xeon's tiny bit more power draw is worth the massive performance improvement.)

[FLY3R]

(Although the Xeon's tiny bit more power draw is worth the massive performance improvement.)

Yes Indeed!

[overcast]

Unfortunately they still don't scale for shit.
But the Opteron still has lower idle and TDP power draw than the Xeon's. :)

(Although the Xeon's tiny bit more power draw is worth the massive performance improvement.)

[DaJMasta]

Too bad they're comparing different Opteron processors.


Intel used an older version.... the 285..... AMD used one of it's newer chips.




Graphs don't lie much, but the headings do. Don't know if I want to blame Intel for using an out of date chip or blame AMD for making the wrong comparison to try and prove their point.

[XooM]

but once you scale beyond 2S it doesn't really matter much at all anyways ;)

[newtekie1]

See AMD, this is what happens when you get too full of yourself and think that since the competition did everything better then you their test must just be biased. You end up proving that their test actually were more forgiving to you then your own and make yourself look like an idiot.

I think the best part is that they made a huge stink about using comparable setups, but the only comparable setup to the Xeon was the seetup intel used. There currently aren't any 3GHz "Next-gen" Opterons on the market. Not that it really matters since the Opterons now fail in comparision to the Woodcrest Xeons in performance too.

[i_am_mustang_man]

they should show data crunching capabilities too, because i bet the optys pwn

[newtekie1]

they should show data crunching capabilities too, because i bet the optys pwn

I am pretty sure clock for clock the Woodcrests(Conroes) outperform the Santa Rosa(Windsor) in essentially everything. Though I base this only on the fact that Conroe outperforms Windsor, and the server processors are the same architectures.

So something tells me it is not in AMDs best interest to compare performance:watt:clock benchmarks.

[jocksteeluk]

gotta give credit to AMD, they didnt believe something so they test it and admit its true and havent tried to discredit it in anyway

[magibeg]

Its all in the fine balance of things... i'm sure in the not too distant future if history continues to repeat itself as it does in every market across the world... amd will eventually unleash their next gen proc and take the lead only to lose it again to intel who will in turn lose it to amd again, and at the end of all this sony's cell chip will dominate from that point on forever (j/k)

[XooM]

gotta give credit to AMD, they didnt believe something so they test it and admit its true and havent tried to discredit it in anyway

where are you getting that from? AMD's tests show opties drawing less power than xeons, intel's show the opposite...

[newtekie1]

Yes, and did you happen to notice that they seem to have neglected to test a 3GHz system?

Apparently though AMD thinks that comparing a 2.6Ghz setup to a 3GHz setup is "systems comparably configured".

[Dippyskoodlez]

Yes, and did you happen to notice that they seem to have neglected to test a 3GHz system?

Apparently though AMD thinks that comparing a 2.6Ghz setup to a 3GHz setup is "systems comparably configured".

whats the problem :confused: they used 2.6ghz optys... high end cpu's against 2.6 AND 3.0ghz xeons..

And the opteron whomped on the 2.6.... :eek: That alone is what they neededto achieve. Letalone the fact that makes them stomp the 3ghz even more...


If you compare a high end systekm to a low end one and it beats it via power consumption, how would that be an unfair comparison, anyways? :confused: Aside from the fact they had a 3ghz compare ;)