The difference betwen 960 at 7xx MHz and 1120 900 MHz is realy small . What would justify the price difference
It seems fillrate and bandwith is still king of the fps.
Well, if the benchmark is bandwidth-limited, it's not surprising…
The difference betwen 960 at 7xx MHz and 1120 900 MHz is realy small . What would justify the price difference
It seems fillrate and bandwith is still king of the fps.
Epeen.The difference betwen 960 at 7xx MHz and 1120 900 MHz is realy small . What would justify the price difference
Asus will be soon launching a line of graphics cards based on AMD's Barts 40nm GPU, and it will include a DirectCU-ed version of the Radeon HD 6850. Named EAH6850 DirectCU/2DIS/1GD5, the incoming card features DirectX 11 support, 800 Stream Processors, and should run cooler than reference models.
Asus' EAH6850 DirectCU also has a slightly overclocked GPU (790 MHz), a 256-bit memory interface, 1GB of GDDR5 VRAM @ 4000 MHz, dual-DVI, HDMI and DisplayPort outputs, and can already be found on pre-order priced at $218.44.
Maybe its just me but at those prices the GTX 460 still very much has a purpose in the market. AMD is still leaving a large gap in their lineup between $140 and $200. I really expected the 6850 to be ~$185.
"The first benchmark is Dirt 2, Ultra DX11 at 1920x1200, 8xAA, 16xAF"
"The benchmark is a a 3-screen [BattleForge] 5760x1080 benchmark, run at DX10.1 mode, comparing HD 6850 Crossfire to GTX 460 SLI. In numbers, the HD 6850 CF is 1.6x 460 1GB SLI and 3.57x 460 768MB SLI."
http://vr-zone.com/articles/amd-radeon-hd-6800-dirt-2-battleforge-benchmarks-leaked/10091.html
BTW on
The 960 (6850) is not CF. It's single.
Nicely directed benchmark though, shows something in the way of improved memory management.
Don't think so the text says otherwise. And my chinese is non-existent, but clearly there are some additional symbols right at "960" compared to the other benchmark with single cards.BTW on
The 960 (6850) is not CF. It's single.
I don't think this is any different than what we sometimes saw with Evergreen vs. Fermi.Nicely directed benchmark though, shows something in the way of improved memory management.
Well according to google translate, the characters, 单卡 mean "single card."Don't think so the text says otherwise. And my chinese is non-existent, but clearly there are some additional symbols right at "960" compared to the other benchmark with single cards.
Ah ok then. Still, this might be a case where SLI is slower than non-SLI due to the memory issues. The chart is totally worthless, even more so since in this case it would only list non-crossfire HD6850 (who says this doesn't have problems with Crossfire due to memory pressure too?) vs. SLI GTX 460.Well according to google translate, the characters, 单卡 mean "single card."
But GTX460 can't run 3 outputs at once without SLI.Ah ok then. Still, this might be a case where SLI is slower than non-SLI due to the memory issues. The chart is totally worthless, even more so since in this case it would only list non-crossfire HD6850 (who says this doesn't have problems with Crossfire due to memory pressure too?) vs. SLI GTX 460.
Right, forgot about the resolution... In any case, the chart just doesn't tell us anything about the performance to expect, unfortunately.But GTX460 can't run 3 outputs at once without SLI.
http://www.pcworld.fr/reactions/xfx-laisse-filtrer-amd-radeon-hd-6870-hd-6850/507365/#commentMSI Radeon HD 6870, 256-bit 1120SP 1GB GDDR5 900MHz 4200MHz DirectX 11.
XFX Radeon HD 6850, 256-bit 960SP 1GB GDDR5 775MHz 4000MHz DirectX 11.
J'ai reçu ça de la part d'un de mes amis qui tiens une boutique.
Found this on PCWorld.fr :
http://www.pcworld.fr/reactions/xfx-laisse-filtrer-amd-radeon-hd-6870-hd-6850/507365/#comment
J'ai reçu ça de la part d'un de mes amis qui tient une boutique. = I got this from a friend of mine who runs a shop.
I guess those are the final specs, then.
That would mean, in comparison with HD5850, HD6870 is:
So, a clear bias in favour of fillrate. Gonna be interesting to see how this works out for real in games...
- GFLOPS - 2016 v 2088 = 97%
- GTexels/s - 50.4 v 52.2 = 97%
- GPixels/s - 28.8 v 23.2 = 124%
- GB/s - 134.4 v 128 = 105%