[ComputerBase.de] Forza 7 Benchmark: Vega has more gasoline in the blood than Pascal

[Yotsugi]

That is if those feature actually work in hardware...

They've already presented some pretty vague numbers for NGG Fast Path in the whitepaper. I assume it kinda-sorta works.
I wish they started the development of consumer drivers earlier than somewhere mid-July.
Products like WX9100 and SSG are nice, and software support for them is even nicer but I don't care about them, I want my Nitro+ 56 and FreeSync monitor.

 

[UglyDuckling]

Go to the 2:06-2:10 mark. Yes there is a difference, but in all likelihood its similar to the same issue involving another scene discussed here in Prey a few months ago, which had to do with lighting and shifting sun/timings. It can look the same as the vega one at different game times. You can even see a hint of that towards the 2:11 mark (vs at 2:06).

Basically all this is desperate fishing for the slightest of hints that AMD has superior IQ to Nvidia, but this conspiracy falls flat in vast majority of games and only in isolated cases we may see one brand differing from the other usually due to game lighting/timing and/or a driver bug/issue that is usually rectified with driver updates. Again, put in perspective, this is rare. What is not rare is the eagerness of some to jump on it milk it for whatever its worth.

I could cook bacon on your tin foil right now.

 

[UglyDuckling]

 

[guachi]

Given the monitors people would actually be playing this game on, it's likely the results would be visually indistinguishable between about of these cards.

Are 4k monitors above 60Hz even available? What's the highest refresh rate on most 1440 monitors?

The takeaway is that all three of these cards can play the game at maximum settings with equal ease.

 

[ZGR]

Go to the 2:06-2:10 mark. Yes there is a difference, but in all likelihood its similar to the same issue involving another scene discussed here in Prey a few months ago, which had to do with lighting and shifting sun/timings. It can look the same as the vega one at different game times. You can even see a hint of that towards the 2:11 mark (vs at 2:06).

Basically all this is desperate fishing for the slightest of hints that AMD has superior IQ to Nvidia, but this conspiracy falls flat in vast majority of games and only in isolated cases we may see one brand differing from the other usually due to game lighting/timing and/or a driver bug/issue that is usually rectified with driver updates. Again, put in perspective, this is rare. What is not rare is the eagerness of some to jump on it milk it for whatever its worth.

Ok, I absolutely see it. That is just the old driver bug from Nvidia... Darn.

I could cook bacon on your tin foil right now.

On rewatching it, it is really obvious that shadows are simply not being drawn on the Nvidia cards. This was discussed to death on the Prey release thread. This was patched a long time ago on subsequent driver releases. No idea why the video is not using updated drivers.

Given the monitors people would actually be playing this game on, it's likely the results would be visually indistinguishable between about of these cards.

Are 4k monitors above 60Hz even available? What's the highest refresh rate on most 1440 monitors?

The takeaway is that all three of these cards can play the game at maximum settings with equal ease.

4k is stuck at 60hz until next year. 4k144hz is due around Q2 2018. 1440p is capped at 165hz but apparently is getting a 200hz bump before the end of the year; maybe even 240hz.

 

[amenx]

Ok, I absolutely see it. That is just the old driver bug from Nvidia... Darn.

Nope, in this case it was just the dumb vid uploader using different settings or something. See developing comments in the vid. The same scene was done back in July when vega released and also compared to 1080/ti:

https://youtu.be/zq8VDDZKp6Q?t=150

No difference.

 

[3DVagabond]

Interesting how in almost all videos, AMD deliver superior image quality, colours.. AA..

In that 2nd video the AO is much better. Almost like it's off on the nVidia cards.

 

[amenx]

In that 2nd video the AO is much better. Almost like it's off on the nVidia cards.

Are you able to point that out in this vid which has many of the same scenes?

https://youtu.be/zq8VDDZKp6Q

 

[amenx]

I could cook bacon on your tin foil right now.

Interesting how in almost all videos, AMD deliver superior image quality, colours.. AA..

You were saying?

 

[3DVagabond]

Are you able to point that out in this vid which has many of the same scenes?

https://youtu.be/zq8VDDZKp6Q

No. But I fail to see what that has to do with the video I mentioned.

 

[UglyDuckling]

You were saying?

I was saying i could cook bacon on your tin foil.

 

[Jackie60]

Go to the 2:06-2:10 mark. NYes there is a difference, but in all likelihood its similar to the same issue involving another scene discussed here in Prey a few months ago, which had to do with lighting and shifting sun/timings. It can look the same as the vega one at different game times. You can even see a hint of that towards the 2:11 mark (vs at 2:06).

Basically all this is desperate fishing for the slightest of hints that AMD has superior IQ to Nvidia, but this conspiracy falls flat in vast majority of games and only in isolated cases we may see one brand differing from the other usually due to game lighting/timing and/or a driver bug/issue that is usually rectified with driver updates. Again, put in perspective, this is rare. What is not rare is the eagerness of some to jump on it milk it for whatever its worth.

Nvidia have cheated knowingly on image quality in the past. Once bitten, twice shy.

 

[amenx]

No. But I fail to see what that has to do with the video I mentioned.

Conversation flow...

Resident Evil 7 - Vega 64 HBBC On vs Off

https://www.youtube.com/watch?v=UALbLTWBQXM

Prey 2017 | Vega 64 LC vs 1080 TI vs 1080

https://www.youtube.com/watch?v=aja0dFdt30A

Interesting how in almost all videos, AMD deliver superior image quality, colours.. AA..

In that 2nd video the AO is much better. Almost like it's off on the nVidia cards.

It was that 2nd video which I addressed here:

Nope, in this case it was just the dumb vid uploader using different settings or something. See developing comments in the vid. The same scene was done back in July when vega released and also compared to 1080/ti:

https://youtu.be/zq8VDDZKp6Q?t=150

No difference.

 

[amenx]

Nvidia have cheated knowingly on image quality in the past. Once bitten, twice shy.

Yes, they have. But rather than diverging much from thread topic (Forza 7 benches), will start separate thread on this.

 

[nurturedhate]

Yes, they have. But rather than diverging much from thread topic (Forza 7 benches), will start separate thread on this.

People already ruined your new thread...

 

[UglyDuckling]

Conversation flow...







It was that 2nd video which I addressed here:

Yes but i'm stubborn beyond all logic, so back to the bacon.

 

[Erenhardt]

Looks like it is CPU-bound. Increasing resolution has little effect on top cards:

Kind of cool you can play it in 4K ultra with midrange GPUs @60fps.

at 4K r9 480=fury=1070! This game loves polaris for some reason
http://pclab.pl/art75785-5.html

 

[Ancalagon44]

@Erenhardt

Looks like that site is using the old Nvidia drivers. With the new ones, Nvidia spanks AMD in this game.

 

[IllogicalGlory]

@Erenhardt

Looks like that site is using the old Nvidia drivers. With the new ones, Nvidia spanks AMD in this game.

http://pclab.pl/art75785-2.html

Those are the latest ones.

 

[Ancalagon44]

Why are their results so different then?

 

[IllogicalGlory]

Why are their results so different then?

I don't think they're testing with the built-in benchmark, or maybe the new AMD drivers also improved performance, or both.

Even with both cards using old drivers, Vega 64 never beat the 1080 Ti at 4K, so the results are different from anything we've seen so far.

 

[richaron]

Why are their results so different then?

Everyone else tests with 8xMSAA and think those results are representative?

 

[Carfax83]

Everyone else tests with 8xMSAA and think those results are representative?

Because MSAA has a very low performance hit in this game. When I had the demo installed, I could run at 4K maxed settings with 8x MSAA and still get a perfect 60 FPS lock. I couldn't go any higher than 60 FPS though because the 60 FPS lock was hard locked. I hear that a newer demo build removed the FPS cap altogether.

Anyway, I don't think these benchmarks from PClab.pl are accurate because the framerates are too low, especially considering that MSAA was barely used. Not only for NVidia, but AMD as well.

Also, HardOCP tested with the latest drivers as well and using actual gameplay, and they used much higher levels of MSAA but got higher performance from both AMD and NVidia.

 

[Muhammed]

Also, HardOCP tested with the latest drivers as well and using actual gameplay, and they used much higher levels of MSAA but got higher performance from both AMD and NVidia.

HOCP is the most representative one, the PClab numbers look crap as hell, RX580 better than 1080 @1080p and 1440p? That's just non sense!

 

[tential]

Is there no MSAA bug with Vega in Forza? How did that ever turn out? I noticed it talked about recently over on other forums, but here it seems to be perfectly fixed. Either that, or there is even more performance there if you reduce the MSAA a little for Vega.