Intel "Harpertown" Xeon vs. AMD "Barcelona" Opteron
by Jason Clark & Ross Whitehead on September 18, 2007 5:00 PM EST- Posted in
- IT Computing
Quest Software Benchmark Factory
We mentioned that the benchmarks we previously used are no longer useful, as we did not have the I/O capacity required to support them. We went looking for alternative benchmarks, and stumbled upon Benchmark Factory from Quest Software. Below is a description of the product and the benchmarks we used in this article.
Benchmark Factory for Databases is a performance and code scalability testing tool that simulates users and transactions on the database and replays a production or synthetic workload in non-production environments. This enables organizations to validate database scalability as user loads increase, application changes are made, and platform changes are implemented. Benchmark Factory is available for Oracle, SQL Server, DB2, Sybase, MySQL and other databases via ODBC and Native connectivity.
Benchmark Factory provides many tests you can run, and has a very nice and customizable metric reporting engine. We decided to run the AS3AP test, and the Scalable Hardware CPU, Reads, and Mixed tests. Here is what Quest's help file says about these tests:
AS3AP
The AS3AP benchmark is an American National Standards Institute (ANSI) Structured Query Language (SQL) relational database benchmark. The AS3AP benchmark provides the following features:
Scalable Hardware
The Scalable Hardware benchmark measures relational database systems. This benchmark is a subset of the AS3AP benchmark and tests the following:
We run three iterations of each load point, and then average the results. We also monitor deviations to ensure they are within an acceptable range. We like to see a max deviation of +/- 3%.
Choosing the contenders
In previous articles, we've been asked to explain why we chose the parts we did for an article. For this article Intel sent us their 3.0 GHz Harpertown CPUs. We requested the 3.0 GHz Clovertown CPUs, which are 120 Watt TDP parts, to allow us to do a clock to clock comparison of Harpertown to Clovertown. We also tried to get Harpertown 2.66 GHz or 2.5 GHz CPUs but none were available. These would have provided us with the closest cost comparison to the Opteron 2350's, but it was not possible. We resourcefully acquired two of AMDs newest Opteron 2350's and we requested the Opteron 2222 3.0 GHz Opteron CPUs, which are the highest clock in the 95 Watt TDP envelope. We did review the results of the Opteron 2224SE 3.2 GHz 119W TDP CPUs but their performance was only marginally better than the 2222's and their performance/watt was consistently lower and thus we concluded of less of interest for this article.
We mentioned that the benchmarks we previously used are no longer useful, as we did not have the I/O capacity required to support them. We went looking for alternative benchmarks, and stumbled upon Benchmark Factory from Quest Software. Below is a description of the product and the benchmarks we used in this article.
Benchmark Factory for Databases is a performance and code scalability testing tool that simulates users and transactions on the database and replays a production or synthetic workload in non-production environments. This enables organizations to validate database scalability as user loads increase, application changes are made, and platform changes are implemented. Benchmark Factory is available for Oracle, SQL Server, DB2, Sybase, MySQL and other databases via ODBC and Native connectivity.
Benchmark Factory provides many tests you can run, and has a very nice and customizable metric reporting engine. We decided to run the AS3AP test, and the Scalable Hardware CPU, Reads, and Mixed tests. Here is what Quest's help file says about these tests:
AS3AP
The AS3AP benchmark is an American National Standards Institute (ANSI) Structured Query Language (SQL) relational database benchmark. The AS3AP benchmark provides the following features:
- Tests database processing power
- Built-in scalability and portability that tests a broad range of database systems
- Minimizes effort in implementing and running benchmark tests
- Provides a uniform metric and straightforward interpretation of benchmark results
Scalable Hardware
The Scalable Hardware benchmark measures relational database systems. This benchmark is a subset of the AS3AP benchmark and tests the following:
- CPU
- Disk
- Network
We run three iterations of each load point, and then average the results. We also monitor deviations to ensure they are within an acceptable range. We like to see a max deviation of +/- 3%.
Choosing the contenders
In previous articles, we've been asked to explain why we chose the parts we did for an article. For this article Intel sent us their 3.0 GHz Harpertown CPUs. We requested the 3.0 GHz Clovertown CPUs, which are 120 Watt TDP parts, to allow us to do a clock to clock comparison of Harpertown to Clovertown. We also tried to get Harpertown 2.66 GHz or 2.5 GHz CPUs but none were available. These would have provided us with the closest cost comparison to the Opteron 2350's, but it was not possible. We resourcefully acquired two of AMDs newest Opteron 2350's and we requested the Opteron 2222 3.0 GHz Opteron CPUs, which are the highest clock in the 95 Watt TDP envelope. We did review the results of the Opteron 2224SE 3.2 GHz 119W TDP CPUs but their performance was only marginally better than the 2222's and their performance/watt was consistently lower and thus we concluded of less of interest for this article.
Facebook
Twitter
RSS
77 Comments
View All Comments
MrKaz - Thursday, September 20, 2007 - link
You bring very valid points! And thanks to the originator of this discussion!But let me spice things a little.
I think you and Anandtech are wrong!
Correct testing would be loading ALL THE MEMORY BANKS WITH RAM!!!
That would be more realistic scenario.
I see Intel praising the technology edge of FBDIMM by allowing to have more RAM on the system, then lets load the Intel system with the maximum RAM they can handle.
Otherwise seams a little biased test.
Showing how Intel systems:
-are energy efficient = use less RAM on them and add more to the AMD system
-can handle much more RAM than AMD = Show how Intel system have lots of memory banks
flyck - Thursday, September 20, 2007 - link
although you are correct when you say there are small errors in the setup, i cant agree with the part about being paid by intel todo...This is an assault which they cannot defend themselves against.
Either way this review would be much more interesting when a 2.5GHz release and low power barcelonas would be available. But that is dependent on AMD itself.
Viditor - Thursday, September 20, 2007 - link
As to that, the low power Barcelonas are available...NewEgg has them in stock already.
http://www.newegg.com/Product/Product.aspx?Item=N8...">NewEgg
flyck - Thursday, September 20, 2007 - link
most hardware site rely on hardware that has been given to them for testpurposes. They wont buy them.Justin Case - Thursday, September 20, 2007 - link
Which is probably one of the reasons why CPUs in some reviews overclock so well, and the ones you buy from retail overclock so poorly.I don't trust any review where the item was supplied by the manufacturer; chances are they cherry-picked the best one they had, to get the best possible review. If the sites can't afford to buy the items they're reviewing, they should simply strike a deal with a retailer, where they get to test the stuff (and return it) in exchange for a sponsored link or something. That way the chances of getting an above-average (or below-average) part are the same as for anyone else.
Justin Case - Wednesday, September 19, 2007 - link
In other words, what you're saying that the Opteron did not have more RAM than the Xeon, so it did not get any benefit from the different memory configuration.Well, that's the "pro-AMD" conspiracy put to rest, no doubt. Thanks.
But you still have 8 DDR2 DIMMs on the Opteron versus 4 FB-DIMMs on the Xeon. As pointed out above, using the same configuration would either reduce the Barcelona system's power consumption (by about 18 watts, if both used 4 DIMMs) or increase the Harpertown system's consumption (by about 40 watts, if both used 8 DIMMs).
In the latter case (which is the likely scenario on a server under high loads - fill it with as much RAM as possible), that would put the Xeon's "performance per watt" below that of the Barcelona system in most of your tests.
And there's still the mystery of why a system that dissipates less heat needs more than twice as many fans. Or was there also a typo on the number of fans in each system? Maybe the number of fans is different but the total number of fan blades is the same, so that's alright? :)
Wirmish - Wednesday, September 19, 2007 - link
The problem is not the number of GB, it's the number of DIMMs.Do you try to convince us that one 8GB DIMM use the same power as eight 1GB DIMMs ?
This is just plain stupid.
Wirmish - Tuesday, September 18, 2007 - link
Same question...AMD..: 8 DIMMs (16 GB) + 7 fans
INTEL: 4 DIMMs (8 GB) + 3 fans
http://www.interfacebus.com/Memory_Module_DDR2_FB_...">LINK
With 8 FBDIMMs the Xeon may consume ~42 watts more !
A standard fan may consume anywhere from 1.6 to 6.0 watts.
Try to use only 4 fans (1 middle-front, 1 top-rear, 2 CPU) with the AMD system.
It will work perfectly and you will save ~15 watts.
1. Add 4 FBDIMM in the Xeon system.
2. Remove three 3.5" fans in the AMD system.
3. Rebench.
4. Update your power consumption and performance/watt graphs.
5. Thank you very much.
Proteusza - Wednesday, September 19, 2007 - link
okay I also think its fishy but I'm playing devils advocate here.if you had to run a netburst server, part of your power goes to cooling, thats part of your total energy requirement. if the AMD system requires more cooling, for whatever reason (no matter how strange that may seem), then like it or not its a part of your server and energy expense.
As for the differing amount of ram, that makes no sense at all. Why halve the amount of memory on the Intel system?
Justin Case - Wednesday, September 19, 2007 - link
Maybe AMD uses inferior knock-off photons, so despite dissipating less heat, it needs more cooling. ;)