differences in Benchmark scores

[TlTimbob]

Hey everybody!
I've been reading a lot of tech sites and the reviews they've been posting about the new graphics adapters from Nvidia and ATI. I've noticed some differences in the results from different sites, and I'm wondering if there may be a reason. Tom's Hardware's scores are consistently higher than those achieved by anandtech and firingsquad. HardOCP and tomshardware have mixed results. I've noticed that there some differences in the system specs that MAY shed some light on this. Tom's and HOCP both use P4 boxes, while anandtech and firingsquad used Athlon64s. Also, Tom's used 2 gigs of memory, while all the other places used 512 megs or 1 gig. Other than that specs were similar, and of course they all used the same version of video drivers. I would also point out that the ATI cards did better on Tom's than on the other sites; that is, they won the benchmarks a lot more often. Hope someone can shed some light on why this may be!

 

[Vapor]

At X-bit the X800XT raped the 6800U, even in CoD. Don't try to start a flamewar saying that THG is ATi biased, EVERYONE had the X800 winning. Test results are only to compare within the REVIEW (not even other reviews on the same site because of driver updates) for the discrepancies you listed.

Now bug off.

Maxtor disgraces the six letters that make Matrox.

 

[Slava]

Now bug off.

What a friendly guy. Oh, I know, I know! His mother taught him not to talk to "strangers" ...

<font color=green>"The creative powers of English morphology are pathetic compared to what we find in other languages." (Steven Pinker, The Language Instinct)</font color=green>

 

[TheRod]

It's normal to a certain extents. I used to have an Excel file that compiled results of various benchmarks from various reviews.

What I have learned, is usually to read at least 2/3 review and then I average results. With averaged results you can clearly see a trend.

You can do the same (I don't have the time to do this for X800/6800 right now) and post your results here.

NOTE : HardOPC have a different approach in their latest hardware review which I really like. Instead of doing same settings comparison, they push the hardware to their effective limit. And the effective limit in games is decent FPS.

--
Lookin' to fill that <font color=blue>GOD</font color=blue> shape hole!

 

[TheGreatGrapeApe]

I agree a bit, I like [H]'s reviews as long as I have a few others I trust to show me head to head at equal settings as well.

My usual sources for a good overall picture are: [H], Digit-Life, Xbit, 3Dcenter, FiringSquad, and Beyond3D.

They all have their strengths, and all do a very thorough job, and when added together, give me what I like to think is a more 'complete' picture.


- You need a licence to buy a gun, but they'll sell anyone a stamp <i>(or internet account)</i> ! - <font color=green>RED </font color=green> <font color=red> GREEN</font color=red> GA to SK

 

[TheRod]

[H] numbers are very useful, because I actually don't care about 300+ FPS in Q3, What I really wnat to know is how these GPU will effectively perform in my PC. When I install a new game, I usually spend at lest 1 hour trying to figure out what is the more effective settings for my GPU. [H] use the same approach, they plug the card and try to figure out how well games will run, this is great!

The only thing that [H] is missing is the same stuff mid-range CPU. I would really like to know if it's worth upgrading to X800/6800 with an Athlon XP 2500+/P4 2.4C or it's a waste of money!

NOTE : I think new GeForce names looks too much like Radeon numbering schemes. We got The Radeon 7/8/9XXX and now we have a GeForce 5/6XXX... Will the next GeForce use 7XXX model names???

--
Lookin' to fill that <font color=blue>GOD</font color=blue> shape hole!