STALKER 2 Performance Benchmark

[mechator]

why is none of the former Quadro cards mixed into the benchmark , like a rtx a5000 it is super awesome for gaming

 

[W1zzard]

why is none of the former Quadro cards mixed into the benchmark , like a rtx a5000 it is super awesome for gaming

This is kinda out of the scope of our audience, but if you send me one I can certainly include it going forward

Performance of RTX A5000 is roughly that of RTX 3080, same GPU, fewer GPU cores

 

[bug]

and yet people still go for the VRAM is META farcry (for god's sake!!) without understanding how the game engine behaves.

Thankful for outlets that aren't halfwits on testing the recomended specs.

The wait is over, but this feels still halfbaked.


Gonna try this one out later, gonna let a 4060ti (8gb) test her out.

If you remember the original, it was in no better shape. Got really good after a series of patches. Though 14 years later and using a 3rd party engine, whether it will be the same story again is anyone's guess.

 

[Nostras]

why is none of the former Quadro cards mixed into the benchmark , like a rtx a5000 it is super awesome for gaming

Because this is extremely niche. Just use TPU's own GPU database and find a GPU near it.

NVIDIA RTX A5000 Specs

NVIDIA GA102, 1695 MHz, 8192 Cores, 256 TMUs, 96 ROPs, 24576 MB GDDR6, 2000 MHz, 384 bit

www.techpowerup.com

Same generation is preferred, which puts it slightly behind the 3080 but significantly ahead of the 3070Ti.
About dead-on with the 4070.

 

[ShiningSapphire]

Why are you not mentioning in game test location?

 

[JustBenching]

If someone is interested, this is what happens when you get into the first "big" city.

Spoiler

 

[W1zzard]

Why are you not mentioning in game test location?

As always we opted for our own custom test scene, which is located in a larger outdoor area. If you stand in a CPU limited area, like the NPC settlements, FPS can be much lower, depending on your CPU power.

 

[Ser_g11]

It would be interesting to see how many processor cores a game can ultimately consume, and is there a limit?

 

[ShiningSapphire]

As always we opted for our own custom test scene, which is located in a larger outdoor area. If you stand in a CPU limited area, like the NPC settlements, FPS can be much lower, depending on your CPU power.

I know but you should describe or show this place, from what point to what point the stalker in that case is moving on the map so I would be able to do such a test also at my platform at home and compare the results.

 

[BrainCruser]

Nooooooooooooooooooope.


Or you know, shaders for specific items were compiled...

Guys get over it, it's UE5 you can't just fix it yourselves. Shader compile stutter are there because devs f'd up and traversal stutter is basically unavoidable in any UE game nowadays.

We aren't talking small stutters of 3 fps for 2 seconds. We are talking stuck at 3 fps for 40-60 seconds, until I look at the ground or sky to free up vram, and go back to 60-70 fps.


Disabling HAGS helped with that.

 

[BSim500]

Game uses Unreal 5. No one should be surprised. It was the same with Unreal 4 & 3.

UE4 had mixed performance, but anyone "remembering UE3 being the same" has a very broken memory as the vast majority of 2010-2014 UE3 games like Bioshock Infinite, Dishonored, Alien Isolation, etc, actually ran very well on High on low-end $100-$150 GPU's of their day (GTX 750Ti / HD 7790 Bonaire, etc) and were an order of magnitude better optimised than today's UE5 ongoing trainwreck...

 

[RootinTootinPootin]

so my makeup test rig

Intel Core i7 12700K @ 5000 MHz - CPU-Z VALIDATOR

[5qkbpg] Validated Dump by rootintootinpootin (2024-11-24 13:47:58) - MB: Asus ROG STRIX Z790-A GAMING WIFI D4 - RAM: 32768 MB

valid.x86.fr

the game defaults to just High with DLSS set to quality

here is a screen nip from in game with native(TAA) vs DLSS enabled

I don't see why HUB is selling that VRAM is META bull crap, when the game defaults to HIGH anyways for the 4060TI, I haven't overclocked the GPU yet nor my PC specs are superior in any way.

Have been playing the game for quite an hour now and there seems to be hovering around 80-100+ FPS with it (default High settings+DLSS Quality) and I must say, even with just High settings, 8GB VRAM is just enough, no stutters/cuts while playing/shooting/free roaming.

 

[Vayra86]

Nooooooooooooooooooope.


Or you know, shaders for specific items were compiled...

Guys get over it, it's UE5 you can't just fix it yourselves. Shader compile stutter are there because devs f'd up and traversal stutter is basically unavoidable in any UE game nowadays.

Its UE5, and you're right, nothing we can do about it, except balance out their lack of investment in a solid engine for the game with a lower purchase price, I'll just wait until this costs 20 bucks now. After all, they've chosen to foot the bill to our minimum required specs for decent performance; I'm already paying top dollar for the suitable hardware to get 60 FPS.

 

[Godrilla]

For reference Metro exodus ran on a 15 year old i7 980xe at 4.3 ghz all cores and 1080ti at extreme settings at 3440x1440p 60 fps.

Its UE5, and you're right, nothing we can do about it, except balance out their lack of investment in a solid engine for the game with a lower purchase price, I'll just wait until this costs 20 bucks now. After all, they've chosen to foot the bill to our minimum required specs for decent performance; I'm already paying top dollar for the suitable hardware to get 60 FPS.

I prebought the game because it was $35. Gamers are paying full price to be a paid beta testers. I believe there is a huge patch coming today that should improve 1000plus bugs or that might have been day one patch. Either way there is one update coming today so hopefully they improve performance and NPCS behavior/ spawn. ( My game had NPC spawn commit suicide after I flexed my new sniper rifle without firing a single shot)
At veteran difficulty.

 

[bug]

It would be interesting to see how many processor cores a game can ultimately consume, and is there a limit?

There is a limit, yes. If you use something like Process Explorer you can see how many threads a process actually spawns. You could give each thread its own CPU and that would be limit. But it would be terribly wasteful, the overwhelming majority of the threads actually do very light processing to the point you can batch dozens on a single CPU and the CPU still won't break a sweat.
(Yes, it is a game of many variables, there is no simple answer.)

 

[neatfeatguy]

Play an hour or two at a time, I'm not having any kind of low FPS hiccups. Just a few minor stutters within the first minute or two after I load into the game. Settings are for 1440p with DLSS off, no frame generation crap, motion blur off and everything set to the default EPIC settings. Game runs pretty smooth for me, averaging around 60fps.

 

[bug]

It usually pays off to fiddle with the settings a bit. More often than not, you will find one or two settings that tax performance more than anything else. And sometimes it's something with minimal visual impact.

 

[rattlehead99]

Is it possible to test games on medium settings with lower-end GPUs? I think that it's quite practical for a lot of people.

 

[harm9963]

Have three games with UE5 ,Hellblade II and Remnant II , and this one , not the best looking out of the three , had zero issues so far !

Side note i did order a new controller , 20 plus year old Xbox 360 died on me , other one died years ago.

Xbox Wireless Controller + USB-C® Cable​

Wow , new controller no issues with dead zone , just perfect !

As always we opted for our own custom test scene, which is located in a larger outdoor area. If you stand in a CPU limited area, like the NPC settlements, FPS can be much lower, depending on your CPU power.

NPC settlements are like cut scene , not that good looking , almost like two different games , also did you play with a controller as well , Xbox 360 is so dated , new wireless controller with cable, got today is awesome , from Microsoft store , they price match ! loaded new driver from device manager .

 

[raptori]

Play an hour or two at a time, I'm not having any kind of low FPS hiccups. Just a few minor stutters within the first minute or two after I load into the game. Settings are for 1440p with DLSS off, no frame generation crap, motion blur off and everything set to the default EPIC settings. Game runs pretty smooth for me, averaging around 60fps.

View attachment 373127

View attachment 373124

View attachment 373125

1440p 60 fps average EPIC settings without DLSS on 3080Ti ! what is your secret ?

 

[W1zzard]

also did you play with a controller as well

Why would I want to be miserable?

 

[neatfeatguy]

1440p 60 fps average EPIC settings without DLSS on 3080Ti ! what is your secret ?

Maybe all the cores/threads with my 5900x....? I even run the 3080Ti at 75% power. I've only just gotten out of the starting area, so I'll have to see how the rest of the game plays.

 

[Vayra86]

Why would I want to be miserable?

Console players will respond with 'I'm not slower with two sticks than you can be with a mouse'

The mysterious wondrious mysteries of perception

 

[JustBenching]

Maybe all the cores/threads with my 5900x....? I even run the 3080Ti at 75% power. I've only just gotten out of the starting area, so I'll have to see how the rest of the game plays.

I'm dropping to 50 in the first village with a 4090 epic and dlss q. So basically 1440p.

 

[ShiningSapphire]

Still no benchmark pass showed. Can u just open the map and draw a line where do you test.