IIRC, the DX12 implementation in UE4 is being done in collaboration with microsoft. So, Gears 4 is probably just a continuation of that.
Preliminary Digital Foundry Analysis of Multiplayer/Horde mode on Xbox One and PC:
PC coverage starts at 6:18. He says the performance is fantastic! Looks like The Coalition successfully integrated DX12 into UE4. Now Epic needs to follow suit and do the same on their end. Heck, they may just use The Coalition's implementation as it seems to work really well so far.
Preliminary Digital Foundry Analysis of Multiplayer/Horde mode on Xbox One and PC:
PC coverage starts at 6:18. He says the performance is fantastic! Looks like The Coalition successfully integrated DX12 into UE4. Now Epic needs to follow suit and do the same on their end. Heck, they may just use The Coalition's implementation as it seems to work really well so far.
but it doesnt really look that good at all.
Well that's your opinion and you're welcome to it. But when evaluating a game's aesthetics, it's important to remove the subjective art from the actual graphics. Gears has always had a slightly cartoonish look to it, and that was all art direction. It's been scaled back considerably for Gears of War 4, but it's still there somewhat. I think that most people that don't like the way the game looks will probably be focusing on the art direction.
But graphically, Gears of War 4 excels, because it runs on the UE4. When you look at the temporal antialiasing (looks clean AF), the PBR (UE4 has some of the best material shaders in the business), particle effects, shadows etcetera and the performance that the engine delivers, then the game really starts to shine.
Also, you do know that most of the footage in that video was from the Xbox One right? And not only that, it was all multiplayer. The MP maps in Gears of War 4 are designed with less focus on graphics (the SP campaign is the opposite) and more on performance because of the competitive aspect.
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 worthy 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 than 1070/980Ti OC, they should 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 to the 1080, 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. We are not in an era where only $100-150 separates an x70 and x80 card like it did during GTX 470 vs. 480, GTX 570 vs. 580 or GTX 670 vs. GTX680 generations. As a result, for $630+ price, 1080 better be smashing 1070 in almost every AAA game by more than 25% given the insane price premium it has.
If we for a second ignore that 1070 and 1080 are mid-range, replace GTX480's price with $630-700 and GTX470's price with $390-410 and suddenly the 1080 looks awful (as in awfully overpriced given its performance) in comparison to historical NV tiers/product positioning. Unlike 480, the 1080 doesn't even have a VRAM advantage over the 1070. We should expect 1080 to shine even more in the latest titles like GoW4 and yet its performance advantage is nothing to talk about. 480 had more VRAM and > 25% advantage at higher resolutions and its price premium was ~ 43%. Since GTX1080 is a far smaller die than the 480 was and has no VRAM advantage, its price premium should be no more than 40% at best. If market price for AIB 1070 is $390-400 now, 1080 should cost no more than $549.
Even if you ignore the aesthetic part of it, it really doesn't match other upcoming UE4 games like paragon, Unreal Tournament and arguably even Lawbreakers IMHO, all of which are multiplayer games.
Personally I think it looks great, and most people that have reviewed it agree with me. The aesthetics match previous Gears of War titles, but the graphics are much improved courtesy of UE4. Also, remember that this is an Xbox One title as well, so the game itself was designed around a much lesser performing console than most gaming PCs..
I think it looks pretty nice as well, I'm just saying that as a 2016 (and close to 2017) title, it doesn't really stand out, in particular not when compared to other notable UE games.
And Paragon (arguably the best looking UE title we have seen so far) is also coming out for PS4, so GoW4 being an Xbox One title is hardly an excuse.
I think it looks pretty nice as well, I'm just saying that as a 2016 (and close to 2017) title, it doesn't really stand out, in particular not when compared to other notable UE games.
And Paragon (arguably the best looking UE title we have seen so far) is also coming out for PS4, so GoW4 being an Xbox One title is hardly an excuse.
I think it looks pretty nice as well, I'm just saying that as a 2016 (and close to 2017) title, it doesn't really stand out, in particular not when compared to other notable UE games.
And Paragon (arguably the best looking UE title we have seen so far) is also coming out for PS4, so GoW4 being an Xbox One title is hardly an excuse.
One thing is for sure, the PC version looks leaps and bounds better than the Xbox One version!
That is a water cooled 1080 running at much higher clocks than the 1070 when both hit about the same clocks in real. You are comparing premium AIB vs standard AIB so clearly not a fair comparison....the difference should be the same 20%.This conveniently ignores all of those gamers with high refresh rate monitors. Anyway, we now have detailed benchmarks for Gears of War 4, so we can finally make a good analysis. GTX 1080 is 29% faster than the GTX 1070 in the 1% low metric..
He didn't test insane settings, but the point is, for high performance gamers, the GTX 1080 will obviously offer a much better experience than the GTX 1070 at 1440p ultra:
No high end card is worthy of its price premium. No matter how many times we tell you, you just don't get it. High end cards are not about price effectiveness. They're about bleeding edge performance!
The GTX 1080 is on average about 20% faster than the GTX 1070. The Titan XP is on average about 20% faster than the GTX 1080. But when you look at the minimums, the gap is often larger as the Gears of War 4 benches above demonstrate.
The larger gap in minimums can make a HUGE difference in whether a game stays above 60 FPS, or drops below it. Case in point:
As demonstrated by the benchmarks above, the GTX 1080 does offer a measurably superior gaming experience, because it allows most games to be played at above 60 FPS at practically all times on max settings at 1440p. In Gears of War 4, a GTX 1080 would be much better for someone with a 1440p 120 or 100Hz monitor than a GTX 1070..
Again, you bring up price effectiveness for a premium product geared towards people that care only about performance. Price is most likely a secondary concern for the people that are buying these products.
I don't even know why you're harping on the GTX 1080. The Titan XP is a far more egregious offender for price effectiveness than the GTX 1080..
What xb1 comparison are you using? Digital Foundry's said that the XB1 changed a lot from the beta to when they tested it the other day.
Obviously I'd expect the PC version to look much better, just wondering what you were comparing to.
That is a water cooled 1080 running at much higher clocks than the 1070 when both hit about the same clocks in real. You are comparing premium AIB vs standard AIB so clearly not a fair comparison....the difference should be the same 20%.
As per TPU the Titan was 30% faster than 1080 and obviously 50% more vram. They haven't put the Titan in the new graphs yet.
OK maybe leaps and bounds is exaggerating, but it definitely looks better, and more importantly, runs better. The 4K textures, and lack of blur bring a clarity to the game that doesn't exist on the Xbox One:
Xbox One vs PC
Xbox One vs PC
Rest of screenshots..
Ya let's see that. 4 1080 AIBs vs 2 1070 AIBs and you are just going by the highest vs highest number and you think that's fair? In my opinion FE vs FE or max OC vs max OC are the most reliable way of comparing. As important as min FPS is it cannot be compared to declare an outright average difference. Average FPS difference has been the standard way to compare GPUs for a while now. If you have a graph of average minimums and frametime then by all means show us till then we have the averages to go with.Did you ignore all the other benches I posted on purpose which show premium AiB vs premium AiB? Asus GTX 1080 is 30% faster than the comparable GTX 1070 in Hitman, and in Rise of the Tomb Raider, the Asus GTX 1080 is 27% faster than the GTX 1070 when you look at the minimum frame rate.
At the time of review, the Titan XP was shown to be 21% faster than the 1080 at 1440p, and 15% faster at 1080p.. At 4K it's 24% faster. Now all that said, I still don't trust TPUs numbers, because they just use the average frame rate, and don't take into account minimums, and even less, frametime, which is perhaps the most important metric..
The Titan XP is probably closer to 30% faster than the GTX 1080 in the real world, just like the GTX 1080 is closer to 30% faster than the GTX 1070 in the real world.
I am a bit confused about this game, does it require a microsoft account? I use a local account on windows 10, and I am not interested in the Microsoft store. Also, if it is so well optimized, where is multi GPU support?
Ya let's see that. 4 1080 AIBs vs 2 1070 AIBs and you are just going by the highest vs highest number and you think that's fair? In my opinion FE vs FE or max OC vs max OC are the most reliable way of comparing. As important as min FPS is it cannot be compared to declare an outright average difference. Average FPS difference has been the standard way to compare GPUs for a while now. If you have a graph of average minimums and frametime then by all means show us till then we have the averages to go with.
I found here a thorough max OC vs max OC comparison for you. The difference at 1440p turns out exactly as 20%. So FE vs FE shows the same difference as well which means whatever variances there are in between AIB vs AIB comparisons are irrelevant to the general comparison of 1070 and 1080. Unlike previous gens none of the AIB cards provide a higher max OC.
http://www.overclockersclub.com/reviews/nvidia_geforcegtx_1070_overclocking/
I used the highest AIB scores for both, but of course since it doesn't suit your views you want to change it.
As old as that review is, it just underscores what I'm saying. The GTX 1070, even when overclocked can barely exceed the 60 FPS threshold, and probably can't sustain it for the heaviest titles.. Look at the results for the Witcher 3 for instance, and the GTX 1070 overclocked gets 56.33, whilst the GTX 1080 overclocked gets 68.66. That 22% difference is huge in the sense that the GTX 1080 can exceed and sustain the 60 FPS threshold, whilst the GTX 1070 cannot.
For all intents and purposes, the GTX 1070 overclocked puts out similar performance to my GTX 980 Ti when it was overclocked. So thats performance I'm already familiar with, and if I had gotten a GTX 1070, it would have been more like a sidegrade than an upgrade.
Anyway, I don't know why you constantly keep projecting your purchase on to me for. Are you seeking validation or something? I don't care what hardware you buy, and you shouldn't care what I buy either..
Enjoy your GTX 1070..
Here's some 4K gameplay on a single GTX 1080 at ultra quality. The frame rate drops below 60 FPS most of the time, but it's still smooth for the most part. One thing is for sure, the PC version looks leaps and bounds better than the Xbox One version!
Makes sense since it's running at 1080p60, while paragon would most likely be 720p60 on xbox.
This game has better PBR and lightmass use as compared to paragon. Of-course Paragon would look better as in that game we don't look from so close as in this one.
I meant camera angle is close to things, player and hence plastic looking thing is more in this one as compared to paragon.