Borderlands 4 Tested: GeForce RTX 5070 Ti & RTX 5080 Performance With and Without DLSS

Introduction

Gearbox Software has released Borderlands 4, and after the developer and 2K Games fast-tracked the release of the game as a show of confidence, PC gamers have sunk their teeth into the title and found it's tragically optimized. Let's see if those claims are true.

68

To see if the claims of Borderlands 4 nuking high-end GPUs are true, I decided to test a GeForce RTX 5070 Ti and a GeForce RTX 5080 across three graphics presets: Badass, High, and Low. I recorded frame data over a 3-minute in-game test using NVIDIA's FrameView, and switched off motion blur. The data has been placed into charts that showcase the performance difference between each of the graphics presets at both 4K and 1440p. Additionally, I did the exact same tests with DLSS on/off to see the performance difference between the upscaling and native.

What were the results? Well, Gearbox Software has clearly developed Borderlands 4 with upscaling as a crutch, as this game is shockingly heavy to run without any upscaling. Even with a GeForce RTX 5080 at 2560 x 1440p at High, not even Badass (the top graphical preset), I wasn't able to hit 60FPS. Yes, that is correct. An RTX 5080 wasn't able to reach 60FPS in an AAA title. That is just one example of what I found with the optimization of Borderlands 4.

Testing Regime & Intent

Each test was run over three minutes and wasn't carried out in a static environment. While many benchmarkers would claim this process jeopardizes the integrity of the test, as I'm not performing the exact same set of actions in every run, I would argue that those results aren't indicative of what a gamer is actually going to experience in-game. No one playing Borderlands 4 will be doing the exact same thing over and over again, which means framerates will vary as players try new things, use different weapons, and so on. This is why each test was carried out over 3 minutes to account for the variability in a general player.

In an attempt to get as close as I can to what an average gamer will experience framerate-wise, I followed the main story for all of my tests. However, I didn't run any tests during cinematics, as that would jeopardize the running average, as they are locked at 30FPS, and I made sure all of my tests included firefights with enemies, which isn't very hard to do in Borderlands 4, as there are enemies everywhere.

Best Deals: Borderlands 4 Super Deluxe - PC Steam [Online Game Code]

Today 7 days ago 30 days ago Buy

Check Price

-

-

CheckCheck Price at Amazon

Check Price

-

-

CheckCheck Price at Amazon

Check Price

-

-

CheckCheck Price at Amazon

Check Price

-

-

CheckCheck Price at Newegg

Check Price

-

-

CheckCheck Price at Amazon

Check Price

-

-

CheckCheck Price at Newegg

Check Price

-

-

CheckCheck Price at PC Case Gear

I also avoided loading into new areas if that is where the story was taking me, as I noticed significant framerate dips when a new region was first approached. Furthermore, I didn't open any menus, such as the inventory, throughout the test. Other than those stipulations, the tests are carried out over completely organic gameplay, meaning I used vehicles, grenades, looted items, spoke to NPCs, aimed down scopes, and caused the same amount of mayhem as any player would in Borderlands 4.

Testing Guidelines & Settings

• Use vehicle at least once
• Engage in combat at least once
• Use graphics presets
• DLSS Quality Preset with Multi Frame Generation x 4

• No menus
• No cinematics/cutscenes
• No new regions
• Motion blur disabled in all tests

If you want to check out the full story for our new testing rig, visit the link below.

Settings Screenshots

4K Badass Native

4K High Native

68

4K Low Native

68

4K Badass DLSS

68

4K High DLSS

68

4K Low DLSS

68

1440p Badass Native

1440p High Native

68

1440p Low Native

68

1440p Badass DLSS

68

1440p High DLSS

68

1440p Low DLSS

68

The same settings were used to test the GeForce RTX 5070 Ti.

4K & 1440p Performance Benchmarks

GeForce RTX 5080 Benchmarks

68

Above shows the performance results I gathered with a GeForce RTX 5080, and boy oh boy, does Borderlands 4 run poorly even with a top-of-the-line PC. At 4K native with the Badass preset, I wasn't even able to hit an average 30FPS, and experienced intense stutters as indicated by my minimum FPS being just 3.2. This was hardly any better at High or Low graphical presets, but what was much better was the minimums being far higher.

At Low, 4K native is actually playable, with an average of 68.2FPS and minimums of around 30FPS. However, this isn't the kind of performance I would expect out of an RTX 5080 paired with an AMD Ryzen 9800X3D, especially given the graphical level of Borderlands 4 in general. This game isn't Cyberpunk 2077 in terms of graphical prowess, and ultimately has no business being this heavy given what is seen on-screen.

As for 1440p, as expected, the performance was better with much higher average FPS, but what is worth pointing out is that at Badass graphics settings, I wasn't even able to hit 60FPS at 1440p. It was only with the High preset that I was able to reach 67FPS.

These native benchmarks highlight that Borderlands 4 isn't optimized at all for native gameplay, and to further illustrate that point, check out the DLSS benchmarks versus the native benchmarks across the same graphics presets.

68

Here we have 4K Native benchmarks versus DLSS Quality with Multi Frame Generation x 4 turned on, which is what Borderlands 4 automatically switches to when DLSS is turned on. As you can see, there is a ridiculous difference in performance in each of the tests, with some of the tests getting 5x the average FPS. Additionally, the minimum FPS was also considerably better, meaning there were far fewer framerate drops to unplayable levels.

68

Moving to 1440p, we can see a similar jump in performance when looking at Native vs DLSS averages and minimums.

This further highlights, in my estimation, that Gearbox Software solely focused the development of the PC version of Borderlands 4 on DLSS and Multi-Frame Generation, and didn't spend much time at all optimizing native.

68

GeForce RTX 5070 Ti Benchmarks

68

68

68

68

Graphics Comparison

Now, the burning question is whether there was a graphical hit when switching to DLSS from native. Honestly, not one that I believe the general gamer would notice if they weren't looking for it. The difference in graphics is negligible compared to other DLSS implementations, and although I noticed a slight blurring or misty effect from the upscaling technology, I don't think it significantly compromised the graphics to warrant not using DLSS.

With that being said, there is quite a substantial difference between each of the graphical presets: Badass, Very High, High, Medium, and Low. The screenshots below showcase a side-by-side comparison of each graphical preset with DLSS on and off.

4K Badass

4K Very High

4K High

4K Medium

4K Low

2560 x 1440p Badass

2560 x 1440p Very High

2560 x 1440p High

2560 x 1440p Medium

2560 x 1440p Low

Final Thoughts

From the above screenshots, you can definitely tell a difference between DLSS on and DLSS being off, but, in some cases, the graphical fidelity of the image has actually improved, not worsened. However, that isn't all aspects. There is definitely some image degradation compared to native in some areas, but generally speaking, it appears DLSS being enabled actually improves the overall image, and doesn't make it worse.

68

There are some downsides of DLSS being enabled, and one of them is input latency. With frame-generation enabled, there is a significant hike in input latency, but like I've mentioned several times, the choice between a slight input latency increase by turning on DLSS or selecting native and barely hitting 60FPS. The choice is a no-brainer.

But I suppose that is wherein the crux of the issue lies. With performance so poor, native gamers are forced to use upscaling technology, as native simply isn't an option. These image comparisons, along with the benchmark results comparing 1440p/4K native versus DLSS, lend to my running theory that the PC version of Borderlands 4 was developed entirely around upscaling technology.

Typically, a developer will adopt upscaling technology after the game has achieved appropriate levels of optimization in native, but when it comes to Borderlands 4, Gearbox Software has seemingly built the game from the ground up around DLSS, leaving native optimization to the bare minimum.

Furthermore, this theory of mine is bolstered by a comment made by Gearbox Software CEO Randy Pitchford, who wrote on X on September 15, "If you are so attached to the idea that DLSS is bad for your game experience because of something you saw or read on the internet, I am sorry. I hope instead you would use the technology designed to give you the experience you want."

While those comments don't directly confirm my theory, it does make it clear that Gearbox is expecting gamers to use upscaling technology if it's available to them. If I'm correct, Borderlands 4 could very well be the first AAA game to be developed entirely around upscaling technology with native optimizations that are so poor that you are left no choice but to use it.