Linux and EM64T; Intel's 64-bit Suggestion

B

burningrave101

[H]F Junkie
Joined
Sep 9, 2003
Messages
11,825
http://www.anandtech.com/linux/showdoc.aspx?i=2158&p=1

Test Setup

As there may have been a little confusion from the last review, the DDR PC-3500 only runs at 400MHz. The Infineon Registered RDIMMs used on the Xeon runs at slightly high latencies. All memory runs in dual channel configurations. We removed 1 CPU for the tests in this benchmark, but since HyperThreading was enabled, we used the SMP kernel. During the second half of the benchmarks, SMP was disabled and the tests were re-run under the single CPU generic kernel. These are both 64-bit CPUs, and so, all benchmarks are run on 64-bit OSes with 64-bit binaries wherever possible.
Click to expand...

Conclusion

Without a doubt, the 3.6GHz Xeon trounces over the Athlon 64 in math-intensive benchmarks. Intel came ahead in every severe benchmark that we could throw at it, particularly during John the Ripper. Even though John uses several different optimizations to generate hashes, in every case, the Athlon chip found itself at least 40% behind. Much of this is likely attributed to the additional math tweaking in the Prescott family core.
Click to expand...

EM64T seems to be doing very well against AMD's x86-64 in a 64-bit environment. I'm looking forward to seeing some more 64-bit benchmarks once we have some 64-bit games and a 64-bit version of Windows XP next year.
 
Nice. It's disappointing that HT enabled seems to hurt/have no benefit more overall than help with most of the benchmarks. But it's good to see that even when the Xeon loses, it's still in the same ballpark. The P4 w/EM64T should perform the same or better with CL2.0/CL2.5 memory since that was only a 1MB L2 Xeon.

Also, the results could be faster with Gentoo or compiling the apps/kernel with the Intel Linux compiler (free non-commercial http://www.intel.com/software/products/noncom/). I might be getting an Intel EM64T processor sooner than I expected to.
 
burningrave101 said:
EM64T seems to be doing very well against AMD's x86-64 in a 64-bit environment. I'm looking forward to seeing some more 64-bit benchmarks once we have some 64-bit games and a 64-bit version of Windows XP next year.
Click to expand...

Don't count on that right away. The reviewer somehow botched the review six ways from Sunday.

http://www.aceshardware.com/forum?read=115093783

Read most of the thread there; reported results for AMD kit under similar configs often get about twice the scores that Kubicki measured. Plus, use of 32-bit vs 64-bit binaries was terribly inconsistent between the two platforms--on MySQL of all things. What was he thinking? :confused:

My best guess is that Kubicki recycled the Athlon64 results from an earlier review, got clumsy, and lost something in the recycling. He probably got in a hurry and got sloppy... :rolleyes:
 
This should have been more of a preview of Intel's 64bit technology rather than a performance bench. Intel's top of the line server chip vs the slowest 939 offering AMD has? I certainly hope the Xeon wins...

Also, read elsewhere on the net, very few people are happy about the method, benches and accuracy of this review.
 
AT has a new review vs the Opteron and the results are very different this time.
 
pxc said:
AT has a new review vs the Opteron and the results are very different this time.
Click to expand...

Exactly.

Top of the line vs. top of the line......... fight!

The reviewer doesn't appear to know much about Linux, as in the first "review" he had a broken Makefile, and his compiler optimizations were incorrect. Clearly he didn't know which ones to use, and when.

In any case, the results are where they should be now.
 
Maybe I should've quoted the guy right above me. It sounded like he didn't know about the Opteron/Xeon matchup. Obviously, in a single processor configuration, the 250 is going to perform identically to the 150.
 
xonik said:
Maybe I should've quoted the guy right above me. It sounded like he didn't know about the Opteron/Xeon matchup. Obviously, in a single processor configuration, the 250 is going to perform identically to the 150.
Click to expand...

Yep my boo boo, I did not read the post right above mine. :eek:
 
You must log in or register to reply here.
Back
Top