There's little to no performance improvement over previous beta

[Snidepeppe]

During the previous beta I did a bunch of testing, unfortunately I did not save the results and the discussion where I posted them is gone, however I remember getting about 80 fps on average at 1080p native lowest settings and the results this time around are:
Average framerate : 80.8 FPS
1% low framerate : 61.5 FPS
0.1% low framerate : 37.7 FPS

Rtx 3080 12gb
Ryzen 7800x3D
32gb ddr5 6000

This is just from the section of the benchmark in the open map, excluding cutscenes since those run much better than gameplay and skew the results; also the fps would drop below 60 when fighting Ray Dau and I would expect the same in the final product.
Regardless of how it compares with the beta this is a terrible running, terrible loocking game.

These are some tests I did some time ago on other games at lowest settings 720p during gameplay in the most demanding area I could find(no built-in benchmark), I was testing ram speeds/timings so they are fully cpu bound:
Cyberpunk
Average: 186.12
1% lows: 115.04
0.1% lows: 71.1

Dragon's dogma 2
Average: 109.82
1% lows: 55
0.1% lows: 35.48

The Witcher 3:
Average: 197.08
1% lows: 104.94
0.1% lows: 57.24

Elden Ring
Average: 149.06
1% lows: 103.88
0.1% lows: 63.26

Baldur's gate 3
Average: 117.82
1% lows: 75.72
0.1% lows: 53.54

Shadow of the tomb rider
Average: 313.34
1% lows: 195.3
0.1% lows: 119.84

And then MH wilds I couldn't even get it to be cpu bound at 720p:
Average framerate : 95.1 FPS
1% low framerate : 72.5 FPS
0.1% low framerate : 47.7 FPS
I did this same comparison during the beta and just like then it berely runs better than DD2, and just like then it seems to not use the GPU to the fullest, my card only pulls 200ish Watts in MH wilds, in other demanding games it pulls around 350 and close to 400 in Cyberpunk with RT.

[RunnyElephant44]

Cadaver původně napsal:

The yalready said this beta is not going to have the perofrmance improvements theyve been working on.

That was the whole point on wanting a second beta they are stupid if it's anything like dragons dogma 2 I'll be mad

[Cadaver]

RunnyElephant44 původně napsal:

Cadaver původně napsal:

The yalready said this beta is not going to have the perofrmance improvements theyve been working on.

That was the whole point on wanting a second beta they are stupid if it's anything like dragons dogma 2 I'll be mad

Maybe it was your reason,but you arent everyone.
The primary reason for this beta is for them to gather data on the how the game plays

[SerpentWave]

You should consider changing the title of this thread to, “ There's little to no PERFORMANCE improvements over previous beta”.

[Snidepeppe]

SerpentWave původně napsal:

You should consider changing the title of this thread to, “ There's little to no PERFORMANCE improvements over previous beta”.

Ah yes you're right

[Snidepeppe]

There may be no performance improvements in the second beta, but they should be there on the benchmark, otherwise what's the point of releasing it, especially as a second separate download. Also I'm not asking to run ultra settings + RT I'm asking for the lowest settings to run well. Given Capcom's track record with improvements over time I have no hope for this one, dragon's dogma 1 was terrible at release and it not improve neither on console nor when it released on PC, same goes for DD2, MH world; to this day in MH rise I can't set some keybinds to certain keys, the game always starts in my second monitor, cutscenes are locked to 30 and more problems still. At least PC features seem to be fine in wilds aside for performance.

[Cadaver]

Snidepeppe původně napsal:

There may be no performance improvements in the second beta, but they should be there on the benchmark,

A few peeps keep mentioning this, but , it doesnt make any sense, the benchmark being a bit more demanding than the actual game is a GOOD thing.
Ideally you want gameplay to be lighter than what you are benching for

[Snidepeppe]

Cadaver původně napsal:

Snidepeppe původně napsal:

There may be no performance improvements in the second beta, but they should be there on the benchmark,

A few peeps keep mentioning this, but , it doesnt make any sense, the benchmark being a bit more demanding than the actual game is a GOOD thing.
Ideally you want gameplay to be lighter than what you are benching for

I don't think the benchmark is more demanding than actual gameplay, there's two cutscenes in there that I had to cut out(one of them is in a tent...) and it is a place with no vegetation, I think fighting a Kushala in the forest area with rain would be far far worst

[Cadaver]

Snidepeppe původně napsal:

Cadaver původně napsal:

A few peeps keep mentioning this, but , it doesnt make any sense, the benchmark being a bit more demanding than the actual game is a GOOD thing.
Ideally you want gameplay to be lighter than what you are benching for

I don't think the benchmark is more demanding than actual gameplay, there's two cutscenes in there that I had to cut out(one of them is in a tent...) and it is a place with no vegetation, I think fighting a Kushala in the forest area with rain would be far far worst

The heaviest part of the benchmark is the cutscene when you enter the hub area.
Fights wont be more demanding than that scene because they aren't as dense in individual entities and objects , which aside from spiking the Draw count massively are all also shadow casters. And thats not even considering the actors running around.

Rain FX are pretty easy to optimize and fighting a single monster in what is essentially a boss arena is much less demanding.
Some effect like a monster attack looking flashy/pretty does not mean it is heavy on processing inherently.
You can do a lot very cheap.

[Snidepeppe]

Cadaver původně napsal:

Snidepeppe původně napsal:

I don't think the benchmark is more demanding than actual gameplay, there's two cutscenes in there that I had to cut out(one of them is in a tent...) and it is a place with no vegetation, I think fighting a Kushala in the forest area with rain would be far far worst

The heaviest part of the benchmark is the cutscene when you enter the hub area.
Fights wont be more demanding than that scene because they aren't as dense in individual entities and objects , which aside from spiking the Draw count massively are all also shadow casters. And thats not even considering the actors running around.

Rain FX are pretty easy to optimize and fighting a single monster in what is essentially a boss arena is much less demanding.
Some effect like a monster attack looking flashy/pretty does not mean it is heavy on processing inherently.
You can do a lot very cheap.

I've kept the hub area in the testing, rain FX might be cheap but reflections are not, but none of this matters, I'm not happy with perfomance as it is, if it will be better in the final game good, otherwise I'll wait to play. Also fight Kushala in MH world in those conditions and then find me a heavier scenario, I don't think there's one.