Charlie at Semiaccurate says: Physics hardware makes Kepler/GK104 fast

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

blackened23

Diamond Member
Jul 26, 2011
8,548
2
0
A 7970 nvidia equivalent for $299 and you guys still complain? Would you rather pay $550 or what? This is good news. Less power draw than a 7970, smokes a 580, beats a 7970 in some games. Sounds like a 8800gt gtx 460 type card that changes the game with an awesome level of value/performance.

ROFL! Are you kidding? For a game that supports physx it will perform well, in everything else it will suck. It will be slower than the 7800. What a joke.

Go for a more traditional card, thanks but no thanks NV. 1 PHYSX game per 8 months isn't gonna cut it, sorry.
 
Last edited:

Keysplayr

Elite Member
Jan 16, 2003
21,209
50
91
You do know how ridiculous that sounds, right? You are talking extremes. Never a happy medium. (No pun intended Happy)
 

Vesku

Diamond Member
Aug 25, 2005
3,743
28
86
Wonder what kind of page hits Charlie is getting for this post backtracking on his short hit-fishing post?

I have yet to see or hear anything that hints that NVIDIA will have a retail 28nm product out before March. Many guesses have been pinned on April/May.
 

96Firebird

Diamond Member
Nov 8, 2010
5,714
316
126
Still, if it's priced like a Pitcairn, and performs at least close to a Pitcairn in the "poor" performance games, then there isn't a big problem, is there?
The problems will arise if it's priced higher than Pitcairn or if it's priced similarly and the performance in the un-optimized games is significantly slower than Pitcairn.

Pitcairn is rumored to be 4 different SKUs, ranging from $200-$300. Who knows which Pitcairn Charlie is talking about, and how much worse it actually is. They are expected to be released this month, so we'll see what their performance is like.
 

MrK6

Diamond Member
Aug 9, 2004
4,458
4
81
This is funny. Amazing how Charlie now speaks the Gospel, when previously his writings of the Fermi launch were an acerbic smear campaign. In the end, Kepler is still no where to be seen, so I guess I'll grab a chair and popcorn in the meantime.
 

blackened23

Diamond Member
Jul 26, 2011
8,548
2
0
You do know how ridiculous that sounds, right? You are talking extremes. Never a happy medium. (No pun intended Happy)

Sorry, I went off on the deep end there. Its just how I read the article, thats what it seemed to convey. Who knows. Its not in production yet so we don't know the entire picture.
 

Keysplayr

Elite Member
Jan 16, 2003
21,209
50
91
This is funny. Amazing how Charlie now speaks the Gospel, when previously his writings of the Fermi launch were an acerbic smear campaign. In the end, Kepler is still no where to be seen, so I guess I'll grab a chair and popcorn in the meantime.

Oh no you don't. Charlie's pulling a fast one on all of us. Don't ever, not even for a minute believe what he says about Nvidia. Good or bad does not matter.
It was a setup to a con. A page hit con. That is all.

Now don't start with the "So you're saying you don't agree with what Charlie says about Kepler?" either. What I AM saying, is that we would all be wise to just ignore Charlie when it comes to his favorite hated topic. Nvidia.
 

AnandThenMan

Diamond Member
Nov 11, 2004
3,979
589
126
It was a setup to a con. A page hit con. That is all.
There is quite a lot of information there, some of it will likely turn out to be true, just as was the case with Fermi. I realize you absolutely loath Charlie and think he is the scum of the earth, but he does his homework. And yes he sensationalizes, that is hardly news.
 

Vesku

Diamond Member
Aug 25, 2005
3,743
28
86
The optimizations part wouldn't surprise me, NVIDIA has been keen on buying/loaning resources for feature locks at least since they bought out 3dfx.
 

blackened23

Diamond Member
Jul 26, 2011
8,548
2
0
The optimizations part wouldn't surprise me, NVIDIA has been keen on buying/loaning resources for feature locks at least since they bought out 3dfx.

I like the part at the end of the article where NV apparently has developers embedded at game studios to sabotage games on competing hardware.

I wonder if thats true.
 

Vesku

Diamond Member
Aug 25, 2005
3,743
28
86
NVIDIA wouldn't word it that way, but they do have a larger better funded developer loaner team than AMD. What are the odds the NVIDIA coding team will NOT try to make their cards perform better than AMD cards, even if that means a hit to overall performance?

Money doesn't even have to change hands directly for this to be worthwhile to a game developer. The longer a game is in development the more valuable additional resources are in regards to getting a salable game out the door.
 

Will Robinson

Golden Member
Dec 19, 2009
1,408
0
0
A 7970 nvidia equivalent for $299 and you guys still complain? Would you rather pay $550 or what? This is good news. Less power draw than a 7970, smokes a 580, beats a 7970 in some games. Sounds like a 8800gt gtx 460 type card that changes the game with an awesome level of value/performance.
Awesome huh!....oh wait...it isn't news.... it's just another unsubstantiated rumor from the rumor factory S/A.:rolleyes:
 

blackened23

Diamond Member
Jul 26, 2011
8,548
2
0
It's not pointless. You do realize that if what Charlie sayz is true we will have a 300$ card that will perform like a GTX580 in all games and smoke every card on the market in certain titles while being a small power efficient chip. Is this bad?

I think you didn't read the article. It will be faster in PHYSX games, and slower than pitcairn in non physx games.

So realistically it will be slower than pitcairn, which is slower than the Cayman XT. Sorry, i'm not seeing an upside to this, I still expected more. It will not perform like a GTX 580.
 

Vesku

Diamond Member
Aug 25, 2005
3,743
28
86
Realistically, Charlie has just covered his bases from slower than the fastest Pitcairn to faster than the 7970. Interesting if the fab spin information is true, Charlie had decent information on Fermi's fab progress.
 

mak360

Member
Jan 23, 2012
130
0
0
Charlie mentions A2; is that the same die shot from the nVidia GPUZ shot flying about on the net. maybe, maybe not.

Charlie mentions PhysX; just a day or so earlier Techeye reported this story http://news.techeye.net/chips/nvidias-kepler-suffers-wobbly-perturbations#ixzz1l4sdeNee

Charlie mentions March/early April; Charlie actually said February before, whereas everyone else said April/May <--(It was also the first date most sites reported until HD79xx had launched)

&#8220;Benchmarks for GK104 shown to SemiAccurate have the card running about 10-20% slower than Tahiti&#8221; Charlie

I think the techeye report may have confirmed what Charlie did not know (PhysX being used) It maybe that Charlie did not know why the big discrepancies between benches and games, which is why he said in his previous article &#8220;wins some & loses some&#8221;. Now he says, it&#8217;s because of PhysX.
 

blackened23

Diamond Member
Jul 26, 2011
8,548
2
0
Charlie mentions A2; is that the same die shot from the nVidia GPUZ shot flying about on the net. maybe, maybe not.

Charlie mentions PhysX; just a day or so earlier Techeye reported this story http://news.techeye.net/chips/nvidias-kepler-suffers-wobbly-perturbations#ixzz1l4sdeNee

Charlie mentions March/early April; Charlie actually said February before, whereas everyone else said April/May <--(It was also the first date most sites reported until HD79xx had launched)

“Benchmarks for GK104 shown to SemiAccurate have the card running about 10-20% slower than Tahiti” Charlie

I think the techeye report may have confirmed what Charlie did not know (PhysX being used) It maybe that Charlie did not know why the big discrepancies between benches and games, which is why he said in his previous article “wins some & loses some”. Now he says, it’s because of PhysX.

In other words,

"he got bamboozled" :D
 

Olikan

Platinum Member
Sep 23, 2011
2,023
275
126
I think you didn't read the article. It will be faster in PHYSX games, and slower than pitcairn in non physx games.

So realistically it will be slower than pitcairn, which is slower than the Cayman XT. Sorry, i'm not seeing an upside to this, I still expected more. It will not perform like a GTX 580.

sudenly this leak makes sence...when you read the small words in the corner
so, maybe it just means that physX won't give a major performance hit....just that

http://wccftech.com/nvidias-kepler-based-gtxgtr-700-series-gpus-performance-compared-geforce-500-series-generation-performance-chart/

EDIT:
doh'p....bf3 don't use physX...forget everyting
 
Last edited:

ViRGE

Elite Member, Moderator Emeritus
Oct 9, 1999
31,516
167
106
What I AM saying, is that we would all be wise to just ignore Charlie when it comes to his favorite hated topic. Nvidia.
Personally, I read Charlie's articles alongside the Sunday Comics. I can always use a good laugh.;)

Seriously though, this article reads like it was written with an automatic essay writer. It's not even stream of consciousness, it's just disjoint and disfigured.