Why do Guru3D & Tech Report Crossfire results differ?

[killtacular]

Hello Peeps,

I'm trying to decide between 2 x 4850 or one 4870. I tend to sell my graphics card every 6-12 months and for that reason feel that the 4870x2 will not be a good buy for me, due to the cost which will be close to £300.

Looking at all the benchies, the 3870x2 is beaten in most games by the single 4870, while costing considerably less. In the UK, 3870x2 is about £230 while the 4870 is £180..... I suppose next year a single mid card ATI may beat the 4870x2 while costing less, and if so, I will be buying that next year...sorry, had to explain my feelings before people say wait for the 4870x2 LOL!

So, back to the question 2 x 4850 or one 4870? I was set on the 2 x 4850 looking at The Tech Report review which shows it beating the 280GTX, and everything else, in almost everything:

http://www.techreport.com/articles.x/14990/11

BUT, the GURU3D says otherwise:

http://www.guru3d.com/article/amd-ati-radeon-hd-4850-review-force-3d--powercolor/19

And now I am confused........

Ta.

 

[fjabad]

you have to look at the hardware they used, totally different

 

[dragonsprayer]

different test systems give different results

that is a game that befits from ati as crysis benfits from nvidia

 

[aevm]

I suggest you decide which games you're going to play for the next 12 months, then find out if they benefit from Crossfire, then buy the 2xHD 4850 if they do, or the single HD4870 if they don't.

Keep in mind that HD 4850CF needs an X38 or X48 motherboard for best results. A P35 would be bad. A P45 would be OK at 1680x1050 but would hurt performance at 1920x1200 or more.
http://www.tweaktown.com/articles/1472/intel_p45_vs_x48_crossfire_performance/index.html

 

[3Ball]

Also, in some cases different reviews will use different testing methods and some may run into driver problems while others may not. There are alot of different factors to consider when looking at different reviews.

Best,

3Ball

 

[Ogdin]

Like mentioned above diff hardware,different drivers,and prob different parts of the games being tested by both sites. Buy whatever you need to play at the resolution you play at.

 

[iluvgillgill]

read more review from different website to get a general/overall performance of the hardware thats being tested.

 

[TheGreatGrapeApe]

The Guru review still has it beating the GTX280 as well, it's not only just different hardware, but diffferet apps, and also different test sections, it'll go up/down depending on all those factors.

Be specific to the games you play or the games that are similar (in workload most importantly, but also developer, and game type) to those you play or plan on playing.

 

[killtacular]

Thanks all for the quick response. I just found it baffling that the very same game "Quake Wars" differs so much in both reviews.

I generally do read all reviews but I have to admit, I am a little biased towards Guru3D when it comes to Graphics reviews. Their write-ups are generally ace.