Its a 180 from winning on all metrics and having a clear winner.
No more hot clocks isnt exclusive to AMD, is it? Also cant you say the same for GCN being more like Fermi? And finally who said that Nvidia has adopted the sweet spot only strategy?
Silent_Buddha
Legend
Not necessarily if it still wins on all metrics compared to its logical competitor.
Regards,
SB
Given the time frames involved, it can't be copying, just convergence.
Charlie likes to poke fun at the uninitiated yet he offered not an ounce of critical thinking in that article. The premise is completely flawed: acceleration of standard PhysX routines will not noticeably increase average frame rates.
The only time PhysX shows that discrepancy is when effects meant to run on the GPU are forced to run on the CPU. The number of titles that even allow that can be counted on one hand.
http://www.behardware.com/articles/804-13/report-amd-radeon-hd-6870-and-6850.html
The only time PhysX shows that discrepancy is when effects meant to run on the GPU are forced to run on the CPU. The number of titles that even allow that can be counted on one hand.
http://www.behardware.com/articles/804-13/report-amd-radeon-hd-6870-and-6850.html
Charlie's original piece compared it to Tahiti, he's now changing his stance by bringing Pitcairn into the picture.
He's doing exactly that in his story.
DarthShader
Regular
You might have to substitute Tahiti for Pitcarn in those claims and they should still work. By the sound of that article, GK104 will be faster or equal on average that Pitcarn in non-PhysX games, while kicking ass in PhysX accelerated games, both being priced at 299$. AMD might have to lower prices, possibly for the 7950 cards too.
Edit: Ninja'd.
So it's unpossible? Nvidia would never go back and patch a game to make it do that?
CD actually says both way but I'm guessing at matching or running slower than Pitcairn in un-Nvidia-optimized games.
$299 would match Pitcairn XT or inbetween XT and Pro in my imaginary product lineup.
Last edited by a moderator:
DavidGraham
Veteran
Ok , so Charlie is being serious here ? his article is filled with some of the worst crap information I have ever seen , dedicated hardware/instructions for Physix , come on !
The only thing I understood from that junk is that Kepler seems to be a strong contender for HD7970 at a much lower price , whether that is ture or not , it remains to be seen !
The only thing I understood from that junk is that Kepler seems to be a strong contender for HD7970 at a much lower price , whether that is ture or not , it remains to be seen !
Patch a game to do what? Make it more CPU limited? How do you propose they do that? And where did nVidia get the power to arbitrarily patch games?
Ye, none of that makes any sense, and I think its clear from the article that not even him fully believe that bullshit.. like its even worse than what he could have dreamt of...
IMO, he is being fed some misinformation.. I cant believe nVIDIA would be as autistic as designing a chip with Pyshx in mind
DavidGraham
Veteran
Yep , that seems to be the case indeed .
And now we know where Charlie obtain his so called "info" from the so called "sources" which are nothing more than some Chinese forum posters !
How dare you claim he's back pedaling? He explicitly writes that he's not!
He's also not a crook...
He's also not a crook...
From what I got after reading that mess is that GK104 is between Pitcairn and Tahiti in terms of performance with some corner cases where it exceeds Tahiti and gets beaten by Pitcairn (how on earth) and has the price/power consumption of Pitcairn XT. Looks like a promising ASIC.
This definitely bodes well for GK110/GK100/GK112 or whatever it's called.
This definitely bodes well for GK110/GK100/GK112 or whatever it's called.
LOL! I was about to...
Doesn't it sound exactly like GF114/110 vs HD6800/6900? GF114 was mostly in the middle of 6870 and 6950, but sometimes fell short of 6870 and sometimes beats 6950
It is similar but more positive for Nvidia because Charlie says in the corner cases it exceeds Tahiti & 580OC meaning it probably is close or ahead of Tahiti XT. Also GF114's power draw exceeded the power draw of Cayman XT and here we have the possibility of GK104 being close to Pitcairn XT.
http://www.bit-tech.net/gaming/pc/2011/07/08/crysis-2-directx-11-patch-analysis/1
Can't get to most of the other links since I'm at work.
DarthShader
Regular
Au contraire, IMO this fits nVidia's mindset very well. What's the most significant competetive advantage nV currently has over AMD? It's PhysX! And if they can't achieve advantages in other areas, like raw rendering power, then what do you do?IMO, he is being fed some misinformation.. I cant believe nVIDIA would be as autistic as designing a chip with Pyshx in mind
Charlie wasn't the first with this rumour anyways: http://news.techeye.net/chips/nvidias-kepler-suffers-wobbly-perturbations
BTW: Do you guys remember the BF3 slide showing massive advantages of Kepler vs. Fermi? : http://wccftech.com/nvidias-kepler-...orce-500-series-generation-performance-chart/
Notice what it says in the lower right corner......... interesting, isn't it?
