Red Dead Redemption 2: PC graphics benchmark review (re-tested)

The_Amazing_X

2020-07-24 14:07

Nice work , thank you. Seems like previus gen is a suprise, Vega 64 just a few fps lower then 1080ti and the gtx 1080 is a bit lower then expected.

#5811863

DannyD

2020-07-24 14:39

Looking forwards to this, thanks!

#5811867

Undying

2020-07-24 14:47

The_Amazing_X:

Nice work , thank you. Seems like previus gen is a suprise, Vega 64 just a few fps lower then 1080ti and the gtx 1080 is a bit lower then expected.

Game favors AMD cards. 5700XT is doing mighty fine for a 400$ gpu.

#5811883

IceVip

2020-07-24 16:00

haste:

Maximum quality tests are pretty much pointless in RDR2. Most of ultra settings are the same shaders with unjustified oversampling, huge performance impact and minimal visual improvements. The only ultra quality setting (except textures of course) I'd definitely keep is the lighting quality. It pushes lights range limits much further and also allows multiple shadowing and doesn't limit moon/sun light in larger towns etc.. it's very demanding at night but it's worth it.

So what exactly are you saying? That Hilbert's tests were straight up pointless because he followed the most common benchmark review format(just max it out)? And that he should make custom settings so that people get even more confused because they think he maxed it out but he actually didn't because he followed his visuals-to-performance instincts?

#5811893

Hilbert Hagedoorn

Administrator

2020-07-24 16:24

The reason for max quality tests being 'standard' can be discussed sure. Yet, I've explained in the article that we'll be using this benchmark with future generation cards in mind, ergo we opt the best quality settings opposed to a mixed setting in the previous revision of the article. It's a terrific Vulkan benchmark for future products. Whether or not you choose max quality is up to you. But that's why I also included some tests with different settings as the game certainly can be tweaked in many many ways. And it just that, the popularity for this game is huge as it's just such a terrific game, and I wanted it included again by adding a Vulkan compatible title into the test suite for graphics cards. Hopefully, future drivers and patches will no longer affect performance.

#5811903

anxious_f0x

2020-07-24 16:53

Thanks for the update Hilbert. Vulkan MultiGPU also works nicely in this game 🙂

#5811912

Jamethe80sman

2020-07-24 17:09

48fps on a 2060 super at 1440p? come on that's just not right my i5 9600k coupled with a 2060s at 1440p on ultra mostly hits 30fps sometimes even lower. Yes it does hit higher fps at less intense parts but for the most its 30fps. Might wait till I get a 3XXX before I go back to this game! I just found it so boring. Loved GTA V but this is just that game reskinned but with less to do.

#5811923

jbscotchman

2020-07-24 17:42

Great article, and its something I've been praising since the game's release. With every patch and driver update my average fps have gone up substantially using a mix of ultra/high and some medium settings. I have done a CPU upgrade since the game came out but that really only helped increase the minimum fps, this game is almost entirely GPU dependent.

Jamethe80sman:

I just found it so boring. Loved GTA V but this is just that game reskinned but with less to do.

Dude, this game has some of the most top notch story driven dialogue, voice acting, and character animations in any game to date in my opinion.

#5811964

HybOj

2020-07-24 19:21

I really miss the comparison with old results, so I can see how the game has developed. Without it, the article somehow misses its point. Still, Im very glad to see such a test like this one, comparing the game in time. Thank you!

#5811966

jbscotchman

2020-07-24 19:28

HybOj:

I really miss the comparison with old results, so I can see how the game has developed. Without it, the article somehow misses its point. Still, Im very glad to see such a test like this one, comparing the game in time. Thank you!

Agreed, but I will give you my personal experience as I have many hundreds of hours in this game since release. Using the same settings @1080p my original average fps was 46, but now its in the low 60's. Like I mentioned I did do a cpu upgrade, AMD FX 8350 to a Zen+ @ 4ghz but only my minimum fps went up, buy a lot actually. However your average fps is entirely based on your gpu which is why this game should not even be considered in CPU benchmarks.

#5811975

jbscotchman

2020-07-24 19:49

Khronikos:

They are definitely not pointless, and this is still how it should be. Only a couple of the settings on Ultra eat loads of frames up, and I agree it would be nice to have a little display of the settings that matter most, but you can find that elsewhere and experiment on your own too. I would say you are wrong that MOST of the ultra settings are worthless. I use a decent amount of them without too many issues. Some of them sure they aren't needed as much and they can eat frames faster.

It reminds me of ubersampling in Witcher 2. Didn't offer any kind of visual enhancement, but totally killed your framerate.

#5811995

vbetts

Moderator

2020-07-24 20:14

Man, Pascal is really showing its limitations in both DX12 and Vulkan in modern titles. Wild to see a card like the 1080 once able to keep up with cards like the RTX series now fall so short, and then turn around to see cards like the Vega 56 and 64 both hold their own in these APIs.

#5812011

Hilbert Hagedoorn

Administrator

2020-07-24 20:45

HybOj:

I really miss the comparison with old results, so I can see how the game has developed.

Though the results between the old test and new test mostly show up in 1920x1080, I could not objectively compare and insert them. The first revision of this benchmark overview was done in-game, for the new one, we moved to the internal benchmark. Secondly, the first article focussed on DX12, this one VULKAN. But as jbscotchman mentions, the differences are to be found in that resolution jumping up 25%~30% Update: I added a comparison of old vs new results on the 1080p page. But again, the results remain subjective from how I look at it, so that is the big disclaimer I am adding here.

#5812016

Astyanax

2020-07-24 21:20

vbetts:

Man, Pascal is really showing its limitations in both DX12 and Vulkan in modern titles. Wild to see a card like the 1080 once able to keep up with cards like the RTX series now fall so short, and then turn around to see cards like the Vega 56 and 64 both hold their own in these APIs.

is it limitations, or is it people expecting too much? theres always going to be something new a card can do that will be taken advantage of in a new games settings.

#5812032

jbscotchman

2020-07-24 21:55

Hilbert Hagedoorn:

Though the results between the old test and new test mostly show up in 1920x1080, I could not objectively compare and insert them. The first revision of this benchmark overview was done in-game, for the new one, we moved to the internal benchmark. Secondly, the first article focussed on DX12, this one VULKAN. But as Jbscortman mentions, the differences are to be found in that resolution jumping up 25%~30% Update: I added a comparison of old vs new results on the 1080p page. But again, the results remain subjective from how I look at it, so that is the big disclaimer I am adding here.

I am jbscotchman, not scortman lol. It's a language barrier thing, I get it. 😛 But just to add further input on the topic when I got a 1440p a a couple months ago there were two games in my mind that I knew I would have to run at 1080p if I wanted acceptable framrates. And those were Control and Red Dead Redemption 2. No matter what cpu, ram amount/speed, etc your GPU will determine your framerate.

#5812049

DannyD

2020-07-24 22:35

jbscotchman:

Agreed, but I will give you my personal experience as I have many hundreds of hours in this game since release. Using the same settings @1080p my original average fps was 46, but now its in the low 60's. Like I mentioned I did do a cpu upgrade, AMD FX 8350 to a Zen+ @ 4ghz but only my minimum fps went up, buy a lot actually. However your average fps is entirely based on your gpu which is why this game should not even be considered in CPU benchmarks.

I'm planning on a 1660S to tide me by till the 3k series so thats very positive news, i'm sticking with my 1080 tele till then too and then switch back to 1440 monitor. This game i've yet to play, probably the one i'm most hyped for, i'll need to get the game soon then.

#5812050

Elder III

2020-07-24 22:37

anxious_f0x:

Thanks for the update Hilbert. Vulkan MultiGPU also works nicely in this game 🙂

MultiGPU is only working with Nvidia gpu(s), or is it working on AMD now as well?

#5812058

jbscotchman

2020-07-24 23:13

DannyD:

I'm planning on a 1660S to tide me by till the 3k series so thats very positive news, i'm sticking with my 1080 tele till then too and then switch back to 1440 monitor. This game i've yet to play, probably the one i'm most hyped for, i'll need to get the game soon then.

Dude, it is such an amazing experience.

#5812061

Hilbert Hagedoorn

Administrator

2020-07-24 23:32

jbscotchman:

I am jbscotchman, not scortman lol. It's a language barrier thing, I get it. 😛 But just to add further input on the topic when I got a 1440p a a couple months ago there were two games in my mind that I knew I would have to run at 1080p if I wanted acceptable framrates. And those were Control and Red Dead Redemption 2. No matter what cpu, ram amount/speed, etc your GPU will determine your framerate.

LOL no clue as to why or how I typed it like that. Corrected, sorry bro 🙂

#5812071

HybOj

2020-07-24 23:58

Hilbert Hagedoorn:

Though the results between the old test and new test mostly show up in 1920x1080, I could not objectively compare and insert them. The first revision of this benchmark overview was done in-game, for the new one, we moved to the internal benchmark. Secondly, the first article focussed on DX12, this one VULKAN. But as jbscotchman mentions, the differences are to be found in that resolution jumping up 25%~30% Update: I added a comparison of old vs new results on the 1080p page. But again, the results remain subjective from how I look at it, so that is the big disclaimer I am adding here.

Thanks a lot, I REALLY appreciate that you did that and the way you interact with the community, this is very special, respect! PS: even if its just indicative, it can be seen that the performance got more in check, for example the 1080ti which behaved sub-par is now more at where it should be etc, I think the comparison is valuable and shows the progress made on the game and drivers. Good to see!