ATi Radeon X1800-kaarten krijgen vuurdoop

Opmerking van iemand van MSN:
Ze zijn erg partijdig, ze pakken erg veel nvidia spellen! Doom 3, Splinter Cell en Far Cry zijn geoptimaliseerd voor nvidia!

ATI's (Chris Hook) first reaction to this article was: " Fiction. I don't believe these numbers were ever run on a 520 ".
A second response just came in:
"This is a complete hoax - done with an OC'd X850, we think. Call our partners yourself and ask if they have R520XTs in house. The numbers aren't even close."

Ik zit nog te wachten op de becnhmarks in Crossfire, dan ownt ati geloof ik makkelijk!

Currently im under NDA so I can not post benchmark results regarding ATI's forthcoming R520 board, but after reading an article today over at HardwareAnalysis I feel its worthwhile giving my views on this.

Hardwareanalysis have been snubbed by ATI and claim to have some results running with the R520 (X1800) which he states have been ran in conjunction with a board partner :

"Normally we would’ve adhered to ATI’s NDA and published results the moment it expired. However this time around ATI decided to just not bother with us as they made it very clear that we did not make the cut, as we have written some articles about CrossFire, R520 and ATI’s financial situation that were apparently not appreciated. Hence we’ve not been briefed by ATI, nor did we sign an NDA, or have been invited to the launch event, so basically they decided to cut us off."

With the help of a board partner we’ve been able to run the same set of benchmarks we’ve previously run on NVIDIA’s GeForce 7800GT and GTX"

Most of you will realise when a website is given hardware by a company like ATI or NVIDIA, testing is done internally by a site reviewer, a member of the reviewing team, I find this article a little disturbing, is this the way hardware sites are going to handle benchmarking in future when "snubbed" by a manufacturer? to have AIB's run tests for public viewing within environments outside our control? There are so many variables involved in this it would certainly not be something I would ever feel comfortable with on any level. It is a website's duty to print their findings to their readers.

I have had some experience with a R520 and I have to say I find the results somewhat suspect. I am not going to get into an "Nvidia V ATI" debacle, the current Nvidia 7800 hardware is first class and has received a heavenly hardware gold award on Driverheaven, the issue at hand are the methods it appears some sites are getting their "results" of unreleased hardware. The results at Hardwareanalysis are scores ive achieved with an overclocked X850XT PE and not a R520 graphics card. I can't help feel with the context of the editorial is it merely a "stab" at ATI for being kept out of the loop.

Speaking with ATI today one of my sources was keen to say "Everyone will see in October that the only board Sander tested was a Ouija board."

My suggestion for all our members reading this is to not get too caught up in "previews" clearly based on bitterness from websites out of the loop but to wait until respected sites like Hardocp, Hexus and ourselves bring figures by testing first hand not via other sources.