 |
|
|
 |
|
|
|
home > news > thoughts on hardware analysis and ati debacle
Thoughts on Hardware Analysis and ATI debacle
Published: 20th September 2005 @ 3:51 AM
Author: Cameron Wilmot
There is no doubt at the moment that ATI is under a lot of pressure from nVidia on all fronts – dual graphics (SLI vs. Crossfire) and next-generation graphics cards (GeForce 7800 GTX vs. Radeon X1800 XT). After owning the graphics card market for a good few years (at least Down Under) since the release of the Radeon 9700 series, ATI had it all their way while nVidia went back to the drawing board and tried to work out what went wrong. Now like ATI were for a few years, nVidia cannot take a wrong step forward and at least from what we’ve seen so far, ATI could be in some trouble. The New York stock exchange backs up these statements with nVidia making big increases in stock prices over the past year or so with ATI going in the opposite direction.
Recently Sander over at European based Hardware Analysis (HA) website posted some apparent benchmarks of the upcoming R520 series (X1800 XT and X1800 Pro). The thing about the benchmarks is that they were (apparently) supplied by a mystery board partner in Taiwan. From what we’ve been told by our friends at ATI Taiwan and various board partners we are real close with in Taiwan, they haven’t had any R520 samples. Not even the ATI office in Taiwan has R520 samples yet - ATI’s HQ in Canada is keeping their next-gen product up their sleeves for as long as possible. We’d never trust benchmarks supplied to us and would never take them and turn them into a full blown article, even if Sander did state that he didn’t run the benchmarks himself – that’s the first mistake he made. On the surface it would seem HA posted the article to either a) generate a lot of hits or b) get back at ATI but we don’t know for sure as everything we have seen cannot be backed up by proof.
Sander could have saved himself a real world of pain by saying who supplied the benchmark results but of course that’s not going to happen in this world or any other. He claimed in a follow-up article that a particular board partner was not pleased with the treatment HA received from ATI but don’t forget this world is all about money and as much as one can get. I can quite safely say no board partner (or employee of a board partner) would ever go behind ATI’s back and provide benchmarks of a product under NDA, it just doesn’t happen – especially not in Chinese culture. Maybe if the board partner was dodgy they’d provide phony benchmark numbers well in ATI’s favor but HA’s benchmarks showed the Radeon X1800XT struggling to make even an impression on nVidia’s GeForce 7800GT. The only company that could have possibly had R520 samples in-house would be Sapphire (the company who makes cards for ATI) and we highly doubt they would have provided HA these benchmarks considering how close Sapphire (who is based in Hong Kong anyway) is with ATI anyway – and we know for a fact Sapphire to this day still doesn’t have R520 samples. Remember everything comes down to bottom lines and Sapphire would not risk their relationship with ATI for a tech site, not at all.
To make things even more interesting, Kyle over at HardOCP quoted an email from Sander to an ATI rep from Europe, “So you're telling me I'm not invited is that it? I feel an ATI column coming up, lets see if we can drop the stock price shall we?. The comment comes because HA was not invited to a European R520 launch event which on the surface seemed to upset Sander. It would seem Sander saw red and decided to attack ATI in the way of posting an article discussing the performance of the upcoming R520 series. Although if you look at the comment another way, it could read: “Screw you ATI, I have details of the performance of your next-gen graphics card and if you aren’t going to include me in on the launch I’ll just post the details now – thank you very much…”
I’m not taking sides of ATI nor HA but I really don’t know what to believe until early next month when websites who are under NDA post their thoughts on the R520 series. Although it would seem safe to say that the benchmarks are inaccurate at best – maybe the benchmarks were legitimately ran by a board partner in Taiwan but possibly with an older version of the R520 or one which is not obsolete and replaced with one with updated clock speeds and so forth. Anyway that’s my two cents on the subject and I suggest everyone wait another couple weeks and find out the facts before looking into this too much.
 Our RSS news feed keeps you updated!
|
 |
| Subscribe to
TweakTown's Newletter! |
|
|
| We will not sell or share your details |
|
|
 |
|
|
 |
|
