AMD: R9xx Speculation

C'mon AMD, the scale starts at .8.

Typical marketing BS. Not the first time and won't be the last. I still remember the V5-5500 graphs vs. whichever nV card it was at the time and literally 2/3 of all charts were like this one showing the V5 as only half the speed of nV when it was arguably very close. Happens.
 
nVidia is apparently sending around eVGA GTX460 1GB FTW (=OC) editions, hoping/asking reviewers to use those against HD6800's instead of reference models

Hm what is with AMD? No OC version?
Here in germany it's very hard to get a reference card. So it would be a little bit stupid to use the reference card when you can't really buy it...
 
nVidia is apparently sending around eVGA GTX460 1GB FTW (=OC) editions, hoping/asking reviewers to use those against HD6800's instead of reference models

I'm surprised nvidia isn't trying to provide guidelines on how AMD cards should be reviewed. It wouldn't surprise me.
 
Hm what is with AMD? No OC version?
Here in germany it's very hard to get a reference card. So it would be a little bit stupid to use the reference card when you can't really buy it...
Obviously those will come later, like it did for the xtx 5870xxx version, etc. Though Barts is supposed to be pin compatible with Cypress, you would think OC versions would be ready from start.
I'm surprised nvidia isn't trying to provide guidelines on how AMD cards should be reviewed. It wouldn't surprise me.
That EVGA card will likely come with guidelines.
 
Just a way to increase the profit margins for AMD, the way I see it. Probably, Nvidia doesn't have anything to do, to keep them doing just that.
 
It's hard to keep track, but was this already posted?

1287495322-650.jpg
 
Nah, anything more would be just fine if they would bring GPU which would breeze trough it.
I'm sure that AMD marketing will change the tone when their GPU solves the small triangle problem. ;)
Marketing yes, but engineering common sense won in this case. Why overkill if you can get where you want with smaller chip and reaching desired performance = bigger profits. Nvidia may brag about advantage in extreme tesselation, which will never see games in the GF100 lifetime, and probably few generations down the road as well.
 
nVidia is apparently sending around eVGA GTX460 1GB FTW (=OC) editions, hoping/asking reviewers to use those against HD6800's instead of reference models
I'm kinda disappointed they aren't releasing a "GTX475" or whatever they would call it full chip GF104. I think it could be even competitive with HD6870 when also clocked around 750Mhz. Guess should wait for official independent benchmarks, but the GTX460 1GB looks like a HD6850 competitor only. (Of course, that would kill GTX465 and GTX470 completely but really GTX470 looks like a bit heavyweight and costly for nvidia as a HD6870 competitor.)
 
I'm kinda disappointed they aren't releasing a "GTX475" or whatever they would call it full chip GF104. I think it could be even competitive with HD6870 when also clocked around 750Mhz. Guess should wait for official independent benchmarks, but the GTX460 1GB looks like a HD6850 competitor only. (Of course, that would kill GTX465 and GTX470 completely but really GTX470 looks like a bit heavyweight and costly for nvidia as a HD6870 competitor.)

They could...
Especially since the 460 reportedly still does better in Heaven vs the 6850 (edited, originally 6870).

But at a 40-60W penalty (if they're gonna clock it higher) they'd have to move into reference-only (or HQ non-ref) territory, and that's the extra surplus cost they'd have to take - when making a card that's gonna sell at say, $270-80 max? (If AMD is targeting $240 for 6870 SRP, and sells extremely sufficiently- pretty much the case as we've seen with Juniper that prices don't even drop)
 
Good to see they finally fixed their AF... though at this point, I'm going to have to see it in a real world application with my own two eyes.

RE other image: You know AMD should consider switching exclusively to red.
 
Back
Top