AMD: R7xx Speculation
- Thread starter Unknown Soldier
- Start date
- Status
...and ATi made it to this very day alive and in well enough shape.
If by 20% market share you mean "good shape"?
No?
AMD is always outsold now 3-1, at the least, in the real world as far as I can tell, by surveys like steam. It might be more like 6-1 or 10-1.
If they keep this up, it wont be long and games will start going Nvidia only (Age of Conan might be the first, I'm hearing it hardly works at all on ATI and the publisher doesn't seem to care). And that will be the end..
In the real world Intel outsold ATI + Nvidia by a ratio of 5:1.
Games only Nvidia? That´s impossible. Games are programed to Direct3d or Opengl and grafic cards run those API´s. If a company makes Bugy games that´s another story. But one game like that will be crap by the early begining.
Another story is that PC game is almost dead. More 2/3 years and only online games, FPS and RTS will left to PC.
Another story is that ATI will increase is market share through 2008. If you knew the grafics history you will see that in 1/2 quarters numbers can reverse very fastly. RV770 + Puma notebooks + 780G/790GX can put market share numbers more equal.
But this is offtopic
So 40, 32, 24 or 16 TMU for RV770?
Last edited by a moderator:
AoC hardly works at all on ATI? Really? Odd, worked OKish as I was testing it prior to release.
Funny, I was looking at the Conan box in EB the other day... maybe I missed it, but Radeon cards weren't listed at all in the requirements section, not even the minimum! (A search online will show what the min. Radeon is, but not the box?) It's a TWIMTBP title, but that is still an outrage and a slap in the face to ATI.
One more thing about perlin noise though: Seems like Radeons did shine in that test ever since:
http://www.pcgameshardware.de/?article_id=624436&page=5
Additionally, it seems not to profit from the more capable TU-circuits in G92.
Leaves us with the question how good a GTX280/260 will perform here.
[/stops playing AoC on his HD2900 Pro]
http://www.pcgameshardware.de/?article_id=624436&page=5
Additionally, it seems not to profit from the more capable TU-circuits in G92.
Leaves us with the question how good a GTX280/260 will perform here.
[/stops playing AoC on his HD2900 Pro]
Perlin is shader bound so when you factor in full shaders + shader clock, you'll notice that G80 lags behind G92 with its much higher shader clock but same # of actual shaders. Of course, in game performance will show that memory bandwidth and other considerations will slow G92 down vs. G80.
What's important isn't the actual Perlin scores relative to other cards from other architectures IMO but rather that it more or less confirms that the RV770 will have 2.01x the shader power of the RV670..
What's important isn't the actual Perlin scores relative to other cards from other architectures IMO but rather that it more or less confirms that the RV770 will have 2.01x the shader power of the RV670..
That's what I said, basically.
aoc at 1680x1050 + 2x fsaa
my radeon 3870x2 is getting 30fps at the worse case and up to 300fps at best case (in door instance with few characters on screen)
this is on a q6600 @ 3.5ghz , 4gigs of ddr 800 ram . My friend who has the same system is getting much worse performanceo n his 8800gtx sli set up.
So i'm not so sure .
all amd needs to do is put out fast gpus . If they can do that they will continue to sell cards . While they might miss out on the halo affect of having hte fastest card out there , many people may be willing to go dual card or x2 if it means getting ther performance of the highst nvidia card for less money
Hey I got this from a NEOGAF thread, not personal experience, so stop yelling at me..sign up and yell at them.
WOW I can't even count the number of nvidia-fanboys over at that kiddie-forum.
If you read through the whole thread you'll see near the end there's a link where they patched and many Nvidia users are actually experiencing similar problems.
I looked at that thread before - people are saying that they are seeing low framerate and GPU utilization by looking at the CCC panel (which reports the level of GPU activity). The GPU has low activity if it is starved of data, so it looks like in the cases that an issue is being reported something the system side of the PCIe interface slowing things down.
That's a pretty big jump from an 8800GTS: +30%! Did that OC bump shader clocks, too?
no-X, thanks for the hints.
I'm really curious why we've seen supposed RV770 leaks, but none for fillrate. Is AMD letting people leak only #s that emphasize their ALU leap? If so, is that b/c it shows them in their best light, or because they want their biggest surprise to remain an equally large bump in texturing power? (That last part sounds like the wishful thinking Slappi's talking about.)
http://publish.it168.com/2008/0610/20080610041602.shtml
4850 3dmark06 scores on a bunch of different settings. No idea if they are real.
4850 3dmark06 scores on a bunch of different settings. No idea if they are real.
- Status
Similar threads
