Review: Assassins Creed: Valhalla graphics performance benchmarks and analysis

True, but it still holds true - memory consumption only goes one way. And then there is the case of lightweight games, where you have plenty of gpu power to downsample, but it reqruires alot more vram.

Like here in cod modern warfare 2 remastered, where im downsampling from 8k



Not gonna be doing that with 10gb vram.

 

Resisting anything and all benchs till I get my 6800xt that have promised myself my once unassailable 2080ti looks so bad on this now and is losing £££s by the Day on Ebay, guess my Wifes 2070 will be getting upgraded, she always gets my cast offs lol

 

Ok than "enjoy" those sweet 40 fps on 3070 @ 4k and tell yourself how good the card works compared to AMD
Your logic is ridiculous. Everyone wants to play at 60fps+. It doesnt matter if 3070 is faster at 4k or 8k or 16k if it cant produce enough frames for the game to be actually playable. 5700xt smoked it where it counts - in the realm of "over 60fps"

 

This isn't a particularly difficult concept to understand. Running the game at higher resolutions puts added load onto the graphics card. In this game, at QHD & 4K, clearly the 3070 pulls ahead of AMD. At 1080p it doesn't - which would indicate that the game is being bottlenecked by something else, or there is some other kind of issue preventing utilization. Which makes your original point "AMD doesnt even need to release new cards, they beat current gen nVidia for 1/2 of price and one generation older tech" a really poor one.

Now that you randomly went from "AMD's entire previous generation is faster than Nvidia at 1/2 the price based one one game, at one resolution, that clearly has some funky utilization going on" to "Well only 60fps matters" I'm still going to stay you're wrong. I think most people, at higher resolutions, would rather turn down a few settings to meet 60fps at their preferred resolution then dropping the res to hit 60fps.

 

None of the reviews seem to want to mention the last few AC games have been running on an engine from 2014, made for PS4 and Xbox One... and when I say "made" I really mean an updated version of Ubisoft's in house engine created in 2007! actually Hardware Unboxed yesterday mentioned it casually in their graphic guide for the game, but didn't bother making it the huge deal that it really is, as it's responsible for all the issues AC games have had in terms of performance.

So we're basically seeing a 10+ year old engine in 2020 with the powerful PC and Consoles we have right now.... the fact that the AC series has always ran poorly has NOTHING to do with the fact it's pushing any sort of boundaries in terms of visual quality, it's just a badly outdated engine that's punching way above its weight class in this new era of hardware.

Also if you think AC games look amazing when you look up close, you need your eyes checked... the textures and pop in alone gives it away as a last gen title not worthy of this new era of PC and Consoles... I mean no Ray Tracing is a dead give away.... Compare Valhalla to Witcher 3 even on the lower settings and you'll see what a proper engine can do for your game, ALSO Witcher 3 is getting a Ray Tracing update for ALL PLATFORMS !

 

Hi, first of all I know that there is no direct comparison between different benchmark scenarios and builds used from different websites/sources but there is no consistency in this case.

Normally you either get a better overall values because the test was perform on a area that was less demanding or simply the build produces better results, but in this case we can see clearly that AMD cards have all more average fps and the Nvidia cards all got fewer across the board.
Something like Vega64 (55.2 -- 60)
RTX3070 (86.2-76)
As a Vega user I cannot complain but this made me a bit confused about the actual performance that I might get when I start this game.



Is there a simple fact that I'm missing for this results?

 

just follow Guru3d reviews

 

lol tpop is so biased its sad. In their review 10700k is still faster than 5900x. That should say something.

 

You would be surprised how well cards can run at 4k with adjusted settings. And this can be done with minimal visual loss. Ppl who are too lazy to experiment with the settings and just use blanket ultra on everything will be those with sub-par experiences with it.

 

Like pretty much every Ubisoft game, performance will get better overall with updates and driver optimizations. I'll just pick this up on sale sometime next year.

 

and what if I add a 15% overclocking performance?

 

Yes, using Hardware Unboxed optimized settings did the trick, at least in the benchmark.

It's been discussed to some length that some games allocate a lot of VRAM but don't necessarily use it. Horizon Zero Dawn is a great example since I played it both on my 2080Ti with 11GB and my current 3080 with 10GB.

When running on my 2080Ti the game ate ~10300MB at 3440x1440. On my 3080 it eats more or less the same amount as in your screenshot ~9600MB but the difference here is that I play at 4K instead of 3440x1440. And the game runs way better on the 3080 (obviously) but I didn't notice any VRAM stutter like I get on my 2GB backup card in most games.

Maybe I'm wrong here but at least it feels like 10GB is OK rn.

 

Holy crap, the 5700xt is kicking ass in this game....

 

It does but overall all amd cards are doing much better. This is what happens when game is optimized for certain hardware. I wonder if we'll see this more and more considering amd is in every console.

Btw, imagine 6800/6900xt benchmarks in this game. Its gonna be a massacre.

 

I mean it's faster in the Guru3D review too.. so lol

Look at the 3090 results, it's back in forth in those games. If you average them, the Intel is actually faster. Toms hardware's 8 game benchmark shows Intel 1% behind average. Is it really that hard to believe that TPUs significantly larger gaming test suite, shows 1-2% faster - without them being biased? I don't think so.

Test conditions (case temps), locations tested, minor settings, ram, etc - can all contribute to difference in results.

Edit: Both testers, tested in game - so that would obviously explain a gap in performance between the two sites.

 

More unoptimized trash by ubisoft

 

Because AMD actually does well or you just being a clown without playing it saying that?
Is a game unoptimized crap when it runs well on Nvidia and not AMD?

 

Wow 5700xt! So new AMD GPUs will rule the top scores? Holy cow...

 

To me, these AMD benchmarks show the kind of gains possible on AMD hardware had they been as involved as Nvidia is in the game design process previously. Glad they are catching up now and making some kickass drivers, actually shows what their hardware is capable of.

Fine wine forever, though bit too late for vega series to get a new lease on life.