EU4 is DX12 compatible now? Last time I checked, it still had DX11 resource management and no multi-engine support implemented.
EU4 is DX12 compatible now? Last time I checked, it still had DX11 resource management and no multi-engine support implemented.According to DualShockers, Gears of War 4 is built from the ground up for DX12, and unlike Quantum Break, it cannot even be backported to DX11 because it leverages so many of DX12's core features, and at a deep level.
I also jacked up settings beyond my rig's near-ultra default and dropped the resolution to 1440p, which still looked gorgeous and rocked a 60fps refresh. I didn't have as much luck pushing some of the settings into the menus' crazy-high "insane" setting, which might be because, according to Raynor, that setting has been offered for future systems and SLI power users: "The game will scale with new hardware that will come out," he said. "You need a really high spec to get in there. It looks awesome, but it's very GPU-hungry." (While I couldn't test Raynor's assertions that the game has been optimized to make the most of SLI graphics card performance, he insisted that users would see substantial boosts if they double-card.)
yeah but I don't think it would be available before final launch of the game.I'm not able to find it in current alpha gameplay videos on youtube
1080 for 1080p......might be recommended reqts.True, but I would imagine that if they are implementing SVOTI in the the game they have tested it to make sure that it can run with acceptable performance even in the heaviest of scenes (although it may very well require a beastly machine)
1080 for 1080p......might be recommended reqts.
10% hit is quite impressive and much better than any of gamework technology.It definetly looks like it could be quite demanding, although the SVOTI feature itself is fairly lightweight (10% hit or so)
If its also coming out on consoles, I'll believe the PC hype when I see it.
In my dream world the next crytek engine will only run on 1080TI's and even if I have to wait 2 years to afford the hardware to play anything using, at least it won't look like a PS3 game.
To think Crysis will be 10 years old next year, still impresses me to this date.
Apparently even a GTX 1080 is not enough for 1440p @ Insane at locked 60 FPS, though the card can do 4K at near Ultra @ 60 FPS.
Performance: Given the very low cost of the game's Anti-Aliasing (0.3 FPS separate Low and Ultra), all players should use Ultra to maximize fidelity,
Nvidia's Gears of War 4 Performance Guide
3 Biggest settings that impact GPU performance on NV cards:
From AT's Pascal review: "In a by-the-numbers comparison then, Pascal does not bring any notable changes in throughput relative to Maxwell. CUDA cores, texture units, PolyMorph Engines, Raster Engines, and ROPs all have identical theoretical throughput-per-clock as compared to Maxwell. So on a clock-for-clock, unit-for-unit basis, Pascal is not any faster on paper."
- OTOH, the game is showing very poor GPU scaling from GTX1070 to 1080 with the latter only 15% faster at 1440p. I expect A LOT better from a $630+ GPU now that 1070 is starting to drop below $400.
Very tricksy. You used the ultra settings benchmark graph to compare the GTX 1070 to the 980 Ti, but when you compare the 1080 to the 1070, you use the insane settings graph which has less of a performance gap. If you use the ultra settings graph instead, the GTX 1080 is nearly 25% faster than the GTX 1070, a larger gap than between the Titan XP and the GTX 1080.
Insane settings are not worth enabling unless you have SLI or when Volta finally hits the deck.
From AT's Pascal review: "In a by-the-numbers comparison then, Pascal does not bring any notable changes in throughput relative to Maxwell. CUDA cores, texture units, PolyMorph Engines, Raster Engines, and ROPs all have identical theoretical throughput-per-clock as compared to Maxwell. So on a clock-for-clock, unit-for-unit basis, Pascal is not any faster on paper."
- 1070 beats 980Ti by 29% at 1080p, and by 24% at 1440p. (1070 = 1920 SPs x 1898 mhz clocks vs. 980Ti = 2816 SPs x 1202 mhz clocks vs. => 7.7% increase only...).
1080p and 1440p ultra and insane settings are all reasonable options. The average difference between 1070 and 1080 turns out 20%. But yes the Titan XP truly is atrocious value which is why even in many no compromise builds you will see a 1080.I'm certain this graph was measured in the context of insane settings rather than ultra. If so, it makes a big difference in the final outcome. Insane settings are not worth enabling unless you have SLI or when Volta finally hits the deck.
To be honest, Ryan really has no method to determine the truth of that comment. Only NVidia knows. I seriously doubt that the improvements for Pascal over Maxwell can be attributed only to clockspeed however.
Very tricksy. You used the ultra settings benchmark graph to compare the GTX 1070 to the 980 Ti, but when you compare the 1080 to the 1070, you use the insane settings graph which has less of a performance gap. If you use the ultra settings graph instead, the GTX 1080 is nearly 25% faster than the GTX 1070, a larger gap than between the Titan XP and the GTX 1080.
Anyway, the overall impressions and previews I've read so far show that this game will be awesome, so I went ahead and preordered it. We should start to see benchmarks tomorrow if I had to guess, as the Ultimate edition will unlock earlier..
To be fair we need min FPS numbers to know if 72 and 90 are practically the same. And they are definitely not exactly the same thing for high refresh rate monitors.That's because at 1440p Ultra, 1070 gets 72 fps and 1080 gets 90 fps. For a 3rd person game, the 90 fps vs. 72 fps isn't a win. The actual experience at 1440p Ultra on a 1070 and 1080 should be identical. Otoh, at 1440p Ultra, the 980Ti is only at 59 fps, which means it'll be dipping below 60 fps a lot more often. Since both the 1070 and 1080 provide identical gaming experience at 1440p Ultra per NV's benchmarks, it's more fitting to compare those 2 cards under Insane settings.
Does 1080 provide the performance that's worth of its 60%+ price premium? 49.9 fps vs. 43.3 fps. Absolutely not! To be able to play this game on Insane settings at 1440p 60 fps a 1080Ti/Titan XP is needed (or 1070 SLI). Once again the x80 series card proves to be a waste of $ (this is the same trend that has remained consistent when comparing 470 vs. 480, 570 vs. 580, 670 vs. 680, 970 vs. 980). As I have stated since May 30, 2016, 1080 sits in no man's land. If someone wants more performance, they buy 1070 SLI, 1080 SLI or Titan XP. In 99% of modern games, it appears that in situations where 1070 cannot provide solid gaming experience, neither can the 1080. I think I saw only 2 games or so (The Division and DE:MD where the 1080 is actually significantly faster).
Maybe they are not, but I expect a card that costs $630-700 to provide a measurably superior gaming experience under the most extreme settings. What else is the point of paying $630-700 for a 1080 over a $390 1070? If you compare Ultra settings, the 1070 is just as good (> 60 fps easily), and 1080 cannot enable higher IQ or 60 fps with Insane, which only proves my point that 1080 is a waste of $.
As demand for 1070 subsides, prices will continue dropping -- such as recent sales of $357, $357 and $348. At these prices, soon it'll be possible to purchase AIB GTX1070 SLI for the price of a single 1080 (certainly a single 1080 FE). In that comparison, the 1080 will get absolutely destroyed as 1070 SLI ~ Titan XP. The more GTX 1070 drops in price and the more 1080 fails to show big benefits over it in modern titles, the less enticing the 1080 looks. We are almost at a point where 1070 SLI costs as much as a GTX1080. Just like 570 SLI, 670 SLI, 970 SLI wiped the floor with 580, 680 and 980, despite 1070's theoretical disadvantage, the same will come to fruition. Therefore, I feel that my point about 1080 showing only a 15% advantage at 1440p Insane in GoW4 is a legit criticism.
I suggest looking at how much resources can get GP100 to each SM, and how big amount of cores is in each SM, in GP104 for example and GM200 chips in the context of GP100. In essence, consumer Pascal GPUs are just Maxwell cards on 16 nm FF+ process.To be honest, Ryan really has no method to determine the truth of that comment. Only NVidia knows. I seriously doubt that the improvements for Pascal over Maxwell can be attributed only to clockspeed however.
Maxwell GPUs performance tanks when it is affected by Asynchronous Compute feature enabled in the game. That is what creates what you see in the benchmarks.Nvidia's Gears of War 4 Performance Guide
3 Biggest settings that impact GPU performance on NV cards:
NV's performance graphs
From AT's Pascal review: "In a by-the-numbers comparison then, Pascal does not bring any notable changes in throughput relative to Maxwell. CUDA cores, texture units, PolyMorph Engines, Raster Engines, and ROPs all have identical theoretical throughput-per-clock as compared to Maxwell. So on a clock-for-clock, unit-for-unit basis, Pascal is not any faster on paper."
- 1070 beats 980Ti by 29% at 1080p, and by 24% at 1440p. (1070 = 1920 SPs x 1898 mhz clocks vs. 980Ti = 2816 SPs x 1202 mhz clocks vs. => 7.7% increase only...).
- Remarkably, the Pascal GTX1060 6GB is only 1.6 fps and 2 fps behind GTX980Ti at 1080p and 1440p, respectively. A $250 next-gen GPU is just a hair behind a 1.5-year-old $650 NV flagship. Looks legit!
- OTOH, the game is showing very poor GPU scaling from GTX1070 to 1080 with the latter only 15% faster at 1440p. I expect A LOT better from a $630+ GPU now that 1070 is starting to drop below $400.
Resolution scaling on a GTX1080
Seeing 1060 ~ 980Ti, fully expecting gimped AMD performance
Per NV, Titan XP is required to hit 60 fps average at 1440p on Insane. GTX1080 achieves 43 fps at 4K on Ultra, not 60 fps.