I havnt read about this on the forums yet, but maybe I missed it. Heres the link http://forums.techpowerup.com/showthread.php?t=53966 Im not so sure this has been discussed properly here. Whats the ramifications? How does Toms writers feel about this? And I welcome ALL writers from all hardware . This is a little disturbing, is it not?

Here is the original story http://www.techpowerup.com/reviews [...] y_9600_GT/ Do the writers acyually know that this happened? And which mobos are effected? or effective? And, how the selling points for this card, which without a PCI-e of over 100MHZ wont show the oclocked ability, and thus make it a lessor card? We as consumers/enthusiasts dont have time to see which mobo has this ability per review. Did Cleeve et al know of this beforehand?

yes the 9600GT is real lol. and now i suppose everyone knows about the clock discrepancies.

Yes, it's been previously addressed here on the forums.

its not that special

What Im wondering is, nVidia didnt tell any of the reviewers of this? or did they?

I just read that article... The only thing I want to know is how 9600GT performance compares to the similarly priced HD3870... I've read reviews and for the most part, they're fairly similar video cards. However, were the reviews based on the Pcie bus speed of 100 for both cards??? If so, then it would be safe to assume that the 9600GT is a better GPU (since you can simply increase the bus to get even better performance). Now if they were tested with an Nvidia motherboard that automatically increased the bus speed, then the 9600GT would clearly be an inferior GPU because even after the "artificial" over clock, the performance is on par with a stock HD3870.

Can someone clear this up for me before I buy a new GPU on wednesday??

Thanks

This is... Interesting. The benchmark results, I suppose, would reflect what the reference system used for its chipset; the nVidia ones would overclock the 9600GT automatically, while the Intel or AMD chipsets used for benchmarking would not.

Still, while it's a novel idea, I'm not yet entirely certain that it would be fully a good idea; this leaves a potential place for errors to arise, and as the article noted, could result, in some cases, of the system not POSTing. And if a piece of hardware, when not defective, won't POST successfully in your machine, it sucks no matter what the benchmarks say.

So here I'm waiting to see if there are any reports of problems with the 9600GT, before I make up my mind on it.

Only checking the motherboards used in each review could clear it up for you.

The one review that I can assume was done with an Nvidia board is the 9600gt in SLI vs. a single 3870x2, where the SLI configuration "won" in two FPS, but the 3870x2 in 3DMark06. Not a comprehensive test, even without Nvidia's tweak.

Nvidia's behavior with this tweak almost reminds me of the Crysis demo drivers that fudged water image quality for benchmarks before the game arrived -- except the boost, if found on all 9xxx series cards on high end Nvidia boards would be legitimate if announced as a board feature. Announced, reviewers could turn it off to provide a level playing field for a 9600gt vs. 3870 comparison.

I'm assuming it would be enabled/disabled in the bios so that if it fails, one can disable it. I'm still amused that Nvidia reps did not know anything about it. It was probably developed at Nvidia's secret headquarters off the California coast, the one guarded by Bond villian extras. IMHO, it's that sneaky a trick.

Should Nvidia's new ad slogan be "Trick or Tweak? The 9600gt does both".

ok, I checked anandtech's article about the HD3870 vs 9600GT. The motherboard they used is the Intel D5400XS (Intel 5400).... Their benchmarks showed very similar results for both GPU's with each of them winning overall fps in certain games (though when the 9600GT won, it was more significant)... Sooo, I guess if you're willing to change the PCIe bus to above 100mhz, the 9600GT should handily out perform the HD3870... On a side note, the 8800GT has dropped prices and the MSI version can be had for around 185 USD on newegg right now... So the obvious choice is the 8800GT since it easily out performs both of the aforementioned GPUs...

My questions:
1. is there any significant features/specs improvements in the newer gpus that would make me consider getting them over the 8800GT??? I don't intend to crossfire, so that's not important... but are the newer GPU features improved significantly enough to justify paying similar prices for lower performance???

2. I've read repeatedly that ati cards provide better video quality where as nvidia cards provide better performance. How much of a difference is the improvement in video quality in ATI cards? Is it worth purchasing the HD3870 over the 8800 GT? I realize it's difficult to quantify something such as video quality, but if you guys say it's a pretty noticeable difference, then I think I'd rather take the performance hit and get the HD3870.