Intel Xeon 3.6 (Nocona) vs. AMD Opteron 250 - Database Test
by Jason Clark & Ross Whitehead on September 13, 2004 12:05 AM EST- Posted in
- IT Computing
Order Entry Stress Test Results
Here is where the architectural differences of the Opteron and Nocona are highlighted. The Opteron 250 managed a 7% gain in performance over the 248, impressive scaling for approximately an 8% clock speed increase. The Nocona story doesn't read as well here as the small-medium work load tests, as saturated bus due to the shared FSB implementation of the Xeon is written all over the enterprise test results. If you look at the 3.2 Prestonia results, the 3.6 Nocona only barely managed a 1% gain in performance, which is really within our deviation between runs. The longer pipeline of the Prescott core, combined with a saturated bus, is not helping the Nocona when the going gets tough. Of course, the Nocona is suffering from 1MB less of L2 Cache than the Prestonia Xeon. It's hard to say how much of a difference that makes at this point.
To give you an idea of the scale of this benchmark, we have graphs of stored procedures calls per second. We decided to focus on Stored Procedures/Second rather than Transactions/Second, since the definition of a Transaction can have a business context or a technical context.
Facebook
Twitter
RSS
44 Comments
View All Comments
RyanVM - Monday, September 13, 2004 - link
#9, why would they be in any rush whatsoever to do that when DDR2 is slower and more expensive clock for clock? Besides, with their on-die memory controller, memory performance is far from a limiting factor for K8 systems (look how little difference there is in performance between single channel and dual channel memory for K8 systems).karlreading - Monday, September 13, 2004 - link
Hmm, seems that the nocona has certainly helped xeons save face against K8. with iAMD64 and 800Mhz FSB they do look much healthier compared to prestona Xeon, raising the perfromance bar and giving the intel fans ( and Dell!! LOL ) the 64 bit option that they always deserved, rather than the expensive, unpopular niche itanic . Considering the implication of having the two CPU's fighting each other for memory bandwidth, something on netburst parts cant get enougth of, the xeons keep pace superbly and it makes one wonder what they would be capable of if each nocona had its own dedicated memory banks and controller!!! Heres one AMD fan hoping opterons @ 2.6Ghz ( opteron 260??) come out quick to keep the chipzilla sweating!!!Viditor - Monday, September 13, 2004 - link
Excellent first (32bit) look, Jason and Ross! Well written, Many thanks!I look forward to the next edition as well.
Jason Clark - Monday, September 13, 2004 - link
Mino, actually AMD sent the tyan and 250's together :) So, they must have some faith in the board since they knew exactly the purpose of it. Aside from that the 250 managed to scale very very well, so I doubt that the board is any issue what so ever... The 64 bit article will be based around 64 bit windows.. not linux. Linux stuff is handled by kris in the linux section.Cheers
mino - Monday, September 13, 2004 - link
by "BEST compiler around" I off course meant by performance, read ability to use the hardware given most efficiently (in case of x86-64 it doesn'nt have to be stable, stability will come by time, these chips will be there for long time).Just a little expl. ;0
mino - Monday, September 13, 2004 - link
hmm, Jason:Please could You explain why with Nocona You do consider it "obviously used the E7520 Lindenhurst server chipset" and with Opteron it is allright to use purely Workstation chipset/board ???
I know there should just minor difference betwen K8W and K8S Pro , but this "minor" rise in K8 performance might have changet this very narrow race !!
Other than That I'm waiting for 64-bit comparison, PLEASE do include 32bit results in there then. I would preffer 32bit Nocona optimized code vs 64bit Nocona optimized vs 32bit K7 optimized(runs better than K8 optim.) vs AMD-64 optimized(compiled with BEST compiler around-even if it will be only beta compiler used) code on Opteron. This should be done on linux with systems compiled from scratch for this test. I know This is much of work, but there are people around that can make such a system up and running in 3 to 4 hours.(one is sitting near me:). So, IF You need help, call some friends of yours. Since such an comparison would be ONE and the BEST ONE around this pitty planet.
good luck,
mino
jrphoenix - Monday, September 13, 2004 - link
#12... Very funny.... I had to go back and read it again. I guess they did fart out the specs, LMAO!!!Jason Clark - Monday, September 13, 2004 - link
A 3.2 prestonia vs. 3.2 nocona could be interesting though, I'll see what we can do there.Jason Clark - Monday, September 13, 2004 - link
We are in planning a 64bit db test, and web test. Not to worry. These types of articles that use real world tests, take time :)Cusqueno - Monday, September 13, 2004 - link
I believe the authors meant "piqued their interest" instead of "peaked their interest". I don't think this is a reference to climbing mountains.