• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Anand's and Tom's 2GHz Reviews: Major differences between the 2?

A

Athlon4all

Diamond Member
I read Tom's 2Ghz article and I am at a loss at why the numbers are so different compared to Anand's. I see 3 major differences in terms of the testing hardware used, in video card, video drivers, and Motherboards.

Anand used SiS 735 based K7S5A, Tom used MSI K7 Master. In P4 Mobo's, Tom used Intel's refrence 850 S478 mobo, Anand used Abit Th7II-RAID. Anand used GeForce3, Tom used GeForce2 Ultra. Finally, Anand used Detonator3, but Tom used Detonator4, but that doesn't matter much because Tom wasn't using a GeForce3.

According to Tom, Athlon 1.4 gets beat in SYSMark 2001 Offfice Apps, that doesn't seem possible, it's not the Athlon's low score that I'm confused at (It's 8% slower, right around the diff between 760 and SiS 735), but what I cannot understand is how Intel's reference 850 board beat Abit TH7II-RAID by 10%.

Q3A no suprises, but another one is AquaMark. I'm gonna quote Tom:


<< The new game AquaNox is already equipped with P4-enhancements, so that Athlon does not stand much of a chance. >>

Ok, now from Tom's view it's no suprise that P4 beats Athlon 1.4 by 3 fps, but what I still cannot understand how does he get that while Anand see's First not only the A 1.4, but the MP 1.2 beat the 2GHz. Did the GF3 that Anand use make this big of a difference?

Finally, CINEMA 4D Raytracing Anand and Tom are about the same, but in iDCT Flask according to Anand the SSE2 algorithm did not make any difference so he sticked with the MMX algorithm, now despite Anand's was a 352 x 288 video and Tom's was 720x576 still would a resoultion increase help the P4 so much and hurt the Athlon?

The bottom line is that I think this is very fishy. The lower Athlon results in Tom's review are understandable, they were about 10% or so less than anand's, that's right in range of how much SiS 735 beats AMD 760. But what I cannot understand at all is how does he get these so much better P4 results, peticularly in SYSMark2001 Office apps? If anything Intel's reference board should be slower than Abit TH7II-RAID, and yet it's faster by as much as 10% in some cases. What do you think? Am I out of my mind?
 
Gotta say I was confused by Anands benchmarks also. In the 1ghz duron review a P4 1.5 with ddr beat a 1.0 Athlon with ddr on most tests (a 500mhz advantage). Now the with the 2Ghz P4 with Rdram, it looses to a 1.4 Athlon on majority (a 600mhz advantage).

Some weird scaling going on here. And the video card differences make it even worse as I thought the GF3 really liked the P4s. 😕
 
Got to love the benchmarks. Kind of hard to compare when it isnt apples to apples on all fronts. What os was Tom using? Anand always uses win2k now.
 
Yes, but he seemed to imply that he had a Socket 478 CPU and P4T is S423. I'm gonna ask him what mobo he used for the 2Ghz.
 
I wouldn't knock the Intel board by automatically assuming it has to be slower than the Asus/Abit crowd. If you're running at default speeds, genuine Intel boards are right up there with the big boys - and sometimes faster. That gets lost on the crowd here at AnandTech where if the board don't overclock at all, it ain't an option.
 
I noticed a was disturbed by the distinct differences between the two reviews myself. The one that really stuck out to me was the Flask benchmark (I wasn't paying that close of attention on the other differences). I dismissed the Flask difference as that Tom was using the special one that was P4 optimized and Anand wasn't until at the end when Anand mentioned that he ran it both ways and the P4 optimized one didn't help at all. That's what I thought was funny. I was really interested when Tom first broke the information about the P4 optimizations of the Flask encoding process because it made a big noise about how easy it was to write programs in P4 optimized code. The difference between regular Flask and P4 optimized was HUGE! Anand should have had a big difference when running the two different modes. The only thing I could think is he is using a different interface to do it or something. I would have to say that at least on the Flask stuff I will stick with Tom's numbers because he spent so much time and effort getting that thing set up.
 
I think Anands score were so much better for the Athlon because he used the MMX iDCT while Tom used the reference iDCT (in the Flask test). But the thing is that nobody uses (well maybe some uninformed people 🙂) the reference iDCT since it offers no better quality then the MMX iDCT, it´s just alot slower (especially on the Athlon).
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top