Discussion in 'Frontpage news' started by Hilbert Hagedoorn, Jul 14, 2020.
Was it really that hard to write "performance" in the title? xD
1080ti is anomaly once per few generations. Its the card that ruined rtx sales.
Is there any way to force DSR in Death Stranding? I would like to test it at 4K, but alas my monitor is 1440p only, and DSR does not seem to work (though it works in any other game).
Check the reviews of the 5500XT 8GB vs it's 4GB version, ty.
The common thing we could see with the rigs of people complaining over the 970's 500MB partition was that they all had system RAM clocked below 2000Mhz.
Yeah, 2.0 is an improvement, 1.0 looks like oilpainting, it's still crap though but better than nothing when even the top end cards don't have the power to do better
https://static.techspot.com/articles-info/1961/bench/CoD_1440p.png - 17% slower.
https://static.techspot.com/articles-info/1961/bench/RDR2_1080p.png - 10% slower.
https://static.techspot.com/articles-info/1961/bench/ACO_1440p.png - 12% slower.
https://static.techspot.com/articles-info/1961/bench/TR_1080p.png - 12% slower.
Minimum framerates are even worse:
https://static.techspot.com/articles-info/1961/bench/1440p_LOW.png - Across basically all those games + more.
I don't know if this game has similar behavior to say CoD, or AC:O. I'm not sure it's worth Hilbert's time testing it and I'm not advocating he does. But I definitely don't agree with the statement that it's not a big difference "like in so many other games" - there are plenty of examples, especially newer titles, where it makes a very large difference, to the point where I'd no longer recommend anyone getting a 4GB card at all anymore. Judging by the VRAM usage in FHD, I'd imagine this game would see a similar hit to the titles above.
Thank you so much Hilbert!
It's pleasing to see, despite it being clear that the VRAM is now most certainly limiting the Fury, that it still has life left in it yet allowing me to hold out until the next generation of GPUs launch. It certainly was the best choice between a 1060 and a 480 when I purchased it.
The Fury did surprisingly well, however indeed 4GB is really a liming factor. Then again, you can lower some quality settings and balance things out a bit better. If only it had 8GB HBM
980ti lives on only becouse it had more vram.
Same for R9 390X and earlier models based on Hawaii with 8GB ... hey at least now you guys know why I have been evangelizing 8GB VRAM for such a long time now.
My 1080Ti and 290x CFX , will never sell them
Hilbert, you haven't re-tested the Radeon vii have you?
As far as I know they never benched the 4GB themselves?
Review here https://www.guru3d.com/articles_pages/msi_radeon_rx_5500_xt_gamingx_8gb_review,1.html and some others the difference is not as much.
But just seen a review where in game like Wolfenstein 2 the increase is 100% while closer to 50% in some of the same games you linked too.
Really weird to to see such huge differences between reviews....
1440p 59 ??? Cool !
No need to imagine
8/8 2019 (AGESA 22.214.171.124) now imagine latest (AGESA combo V2 PI 126.96.36.199) another 3-5% performance uplift for this games
3700x - gtx 970
witcher 3 1% low 64 fps
crysis 3 1% low 91 fps
Assassin's Creed: Origins 1% low 51 fps
Shadow of the Tomb Raider 1% low 51 fps
Watch Dogs 2 1% low 62 fps
Far Cry 5 1% low 63 fps
perfect combo for VERY 1 % lows
>So if you have an RTX card, quality mode ON, end of story (as the increase in your FPS is ginormous).
what? is that statement correct? FPS increase when DLSS Quality fps are lower on chart?
Correct, they would need to use DLSS set to performance to get a performance increase.
Not sure with Nvidia but with AMD for games where the higher resolution doesn't show, you have to set the desktop resolution first to the desired one and then it should appear in the game as well.
Nice work on this article HH. I wasn't so sure about the one on F1 2020 because you stated it's not particularly challenging for CPU or GPU. That made me wonder why you even bothered, to be honest. Using this game for this kind of test makes more sense because it seems to be able to give a worthy challenge to GPUs.
This is the only reason I upgraded my Fury X to a Vega 56. The Vram. The performance of the Fury X was still really good. But 4gb isnt enough for 3440x1440.
I did two quick tests myself. Basically not benchmarks per se. Just a 60fps capped gameplay of the first few minutes, to see how the 2500k and GTX970 fair. Not bad for a 2011 cpu with a 2014 gpu. (non monetized/sponsored channel-I am not clickbaiting anyone)
Also the same on 8600k+GTX 1070, but this time uncapped for the gaming session. The cutscenes are capped anyway.
Also does anyone know why my 7950 does not run this game? It gives an error saying to check the card and the drivers. However I have run DX12 games successfully in the past. So what gives?