This benchmark seems fishy to me. It seems like this person was paid by AMD. Since the SQL database is mostly an I/O benchmark, I wonder why the author would say that Opteron is faster than Woodcrest. Clearly a faster hard drive will give you better results more than a better CPU. I don't even think these tests stress the CPU that much. What were the loads? Was CPU maxed at 100%? If it's not than Woodcrest is at a disadvantage because FB-DIMM is slower than DDR.
Surely, there are better tests to determine CPU speed. Anandtech did a very good review of Woodcrest vs Opteron.
its totally correct
only problem all of the tests were run with netburst xeons.
i am fairly sure new C2D based xeons are what i would do the tests with
you can buy them at the store
According to the testers, the Xeon was indeed Woodcrest (5160)
Here's a good Opteron vs. Woodcrest benchmark. Woodcrest is indeed a good deal faster than Opteron across a wide variety of workloads.
Tech Report is pretty consistent but they need to follow up with a more level comparison. Tossing in the dual Clovertowns kinda smacked of Intel bias when benched against the 4-core of 2218's. For a better head-to-head Opteron/Clovertown bench the SuperMicro H8QM8-2+-O Quad Socket F would be a better choice for an **8 on 8** core vs core comparison. And the Opty rig with 4-way 8218's would be a closer price comparison with the Clovertown rig . . . ANYWAY . . . .
I was also a little disappointed that they benched the 3.0GHz Woodcrests against the 2.6GHz 2218's - especially when there is a $1,000 price difference between the cpu/mobo's. The Opty 2220's @ 2.8GHz would still be $700 cheaper than the Woodcrest rig.
BTW did you check out the Quad FX - Core2 comparison (and cross reference the Xeon - Opty benchies) ??? Kinda raised an eyebrow across the board . . .
Bet the Intel / AMD boards of directors don't want to see too many of these benchies . . .
Xeon (Woodcrest) 5160 and the Opteron 8222 server; clocked both at 3.0GHz
Conclusion:
The test results presented in this document appear to support the position that large and complex applications, like Relational Database Management Systems, run as fast or faster on an Opteron based server as on a Xeon (Woodcrest) server with a virtually identical configuration and processor clock speed.
This benchmark seems fishy to me. It seems like this person was paid by AMD. Since the SQL database is mostly an I/O benchmark, I wonder why the author would say that Opteron is faster than Woodcrest. Clearly a faster hard drive will give you better results more than a better CPU. I don't even think these tests stress the CPU that much. What were the loads? Was CPU maxed at 100%? If it's not than Woodcrest is at a disadvantage because FB-DIMM is slower than DDR.
It's not fishy. SpecFP_Rate benchmarks show single to low double digit performance gains for Opteron vs Woodcrest at same core. Server/workstation is where Opteron shows its prowess with its platform.
You are about to answer a thread that has been inactive for more than 6 months. If you still wish to proceed, please ensure that your posting is original and does not duplicate or overlap any prior responses to this thread.