Borderlands 4 GPU Benchmark: Over 40 GPUs Tested

[PierreJG]

45k+ for a 5090 in South Africa and this is the current results... It will get better with optimizations. I still feel that BL 3 is not correctly optimized. Look at borderlands 1 and 2, those games ran perfect even on older hardware. Obviously can't compare the old vs new but borderlands ran on a freaking gtx9800 gpu with 1gb vram . Just saying it should at least get proper fps on 1080p entry level cards.

 

[Jarenien]

No one should give a crap about native performance, if you're not going to show performance using 4x MFG and all that DLSS4 has to offer, then this is a pointless post, because anybody who isn't a complete and utter *****, will be using those technologies. Not using them would be like doing advanced math with a scientific calculator but not using any of the stuff that makes it so much better than a basic calculator. Sure you could do it, but it would suck, just like gaming without DLSS tech, so just stfu already and just use it you morons.

 

[alexnode]

That's pretty terrible, I think UE5 needs to improve. Or just the artists and developers. My impression of working with unreal is that loads of artists and devs use shortcuts but at the end of the day you need boring hard work for LODs, unwrapping, shader optimisation... deep C++ scripts . Things can be super intense in a tech demo but when you have giant worlds it breaks the bank. IT could be just as simple as redesigning the cell shading rendering style ... could it be that nanite and cell shading is a bit of a nightmare ... I can envisage all these outlines streaming in and out of nanite levels of detail ... I am going to buy it if the make it nice for 3080+ at 4k 60 fps ...

 

[PierreJG]

After buying my RX 6600XT 2 weeks ago. I eventually got around to test some gaming with it. I was suprised to get 90fsp on high settings 1080p. It is not godlike but with a game that bad optimized I was suprised. I am guessing that BL4 will be worst. I won't even bother as I didn't even like 3. They took everything that BL 1 and 2 had to offer and just did a cut down version. I call BL3 BL 0.5 as it is crap compared to the first 2.

 

[PierreJG]

No one should give a crap about native performance, if you're not going to show performance using 4x MFG and all that DLSS4 has to offer, then this is a pointless post, because anybody who isn't a complete and utter *****, will be using those technologies. Not using them would be like doing advanced math with a scientific calculator but not using any of the stuff that makes it so much better than a basic calculator. Sure you could do it, but it would suck, just like gaming without DLSS tech, so just stfu already and just use it you morons.

You actually missed the point of the benchmark. Yes, they can use DLSS/FSR etc but they tested real performance which upscaling doesn't provide. Obviously with upscaling you will get a lot higher FPS but this represent real performance with different GPUs as DLSS/FSR works differently. Native vs Native, you can compare the GPUs directly with upscaling you cannot. So basically what you suggest is that they Test an RX6600 XT vs an RTX 3060 with upscaling when the 3060 DLSS is better than FSR of the RX 6600 XT

 

[godrilla]

Update Borderlands 4 just got a got a patch that significantly improves performance seeing 45% delta gains. FYI

 

[Tam 212]

Update Borderlands 4 just got a got a patch that significantly improves performance seeing 45% delta gains. FYI

Somebody must have "shown" Randy how to code a UE5 game for better performance...

At this point, not sure the optimizations it should have had at launch nor the free bounty pack 1 and the raid boss this week will move the needle at all - at least as far as player count goes.
SteamDB Borderlands 4 Charts