Stellar Blade feature-4

Stellar Blade Benchmarks & PC Performance Analysis

And the time has finally come. Sony has just released Stellar Blade on PC. Powered by Unreal Engine 4, it’s time now to benchmark it and examine its performance on PC.

For our benchmarks, we used an AMD Ryzen 9 7950X3D, 32GB of DDR5 at 6000Mhz, AMD’s Radeon RX 6900XT, RX 7900XTX, RX 9070XT, as well as NVIDIA’s RTX 2080Ti, RTX 3080, RTX 4090, RTX 5080, and RTX 5090. We also used Windows 10 64-bit, the GeForce 576.66, and the Radeon Adrenalin Edition 25.5.2 drivers.

Shift Up has added a respectable number of PC graphics settings to tweak. PC gamers can adjust the quality of Shadows, Textures, Lighting, Characters and more. The demo also comes with the 4K Texture Pack that is exclusive to the PC version. Plus, there is support for both NVIDIA DLSS 4 and AMD FSR 3.0. Sadly, though, there is no support for Intel XeSS.

Stellar Blade PC graphics settings-1Stellar Blade PC graphics settings-2Stellar Blade PC graphics settings-3 Stellar Blade PC graphics settings-4Stellar Blade PC graphics settings-5Stellar Blade PC graphics settings-6

Stellar Blade does not have a built-in benchmark tool. So, for our benchmarks, we used the Prologue/Beach area. This appeared to be the most demanding area early in the game. As such, it should give us a pretty good idea of how the rest of it runs. Also, for those interested, we have a separate article for our DLSS 4 benchmarks. So, be sure to check it out.

Stellar Blade benchmark scene

At 1080p, we were significantly bottlenecked by our RAM or CPU. Thus, I’ve decided to test different CPU configurations to see how the game scales on various CPU configurations.

So, at 1080p/Max Settings, we were able to run Stellar Blade with over 70FPS, even on a dual-core configuration. With 4 cores/threads, we were able to raise our minimum framerate to 119FPS, and our average framerate to 152FPS. Then, we hit a CPU/RAM wall with more than six CPU cores/threads. Despite these bottlenecks, the good news is that you will be able to game with over 60FPS, even if you don’t own a high-end CPU.

Stellar Blade final benchmarks-1

At 1080p/Max Settings, all our GPUs were able to provide high framerates. Even the NVIDIA RTX 2080Ti was able to push over 100FPS. Thus, we can safely say that Stellar Blade will run smoothly on a wide range of PCs. Moreover, our AMD Radeon GPUs were faster than our NVIDIA GPUs, especially when it came to the minimum framerates.

Stellar Blade final benchmarks-2

At 1440p/Max Settings, all of our GPUs were, once again, able to offer a smooth gaming experience. And yes, even at 1440p, the NVIDIA RTX 2080Ti was able to achieve framerates higher than 70FPS. I don’t remember the last time that GPU was able to provide such a smooth gaming experience at 1440p.

Stellar Blade final benchmarks-3

As for Native 4K/Max Settings, our top five GPUs were able to provide framerates over 60FPS at all times. The NVIDIA RTX 3080 and AMD Radeon RX 6900XT came close to a 60FPS experience, though they could drop to the 50s. If you want to game at 4K/Max with those GPUs, I suggest using DLSS or FSR.

Stellar Blade final benchmarks-4

Graphics-wise, Stellar Blade has a great art style and it is pleasing to the eye. Tech-wise, though, it does not bring anything new to the table. For the PC version, Shift Up added some 4K textures that can improve the game’s overall image. For its amazing performance, the game looks great. Just don’t expect it to knock your socks off. It looks great, but not mind-blowing.

Before closing, I want to say that I didn’t run into any big stuttering problems while playing. There were a few shader compilation stutters. However, they weren’t as bad as in some other recent games. Most of the time, the game ran smoothly on PC, even with an unlocked framerate.

All in all, Stellar Blade is a polished PC game. It does not require a high-end PC, meaning that it will run great on wide range of PC systems. It also looks great, and the KB&M implementation feels solid. Plus, it supports some PC-only features, like DLSS and FSR. And although it uses Denuvo, it does not suffer from any performance issues. With the way the game runs, it’s a bit disappointing we did not get any Ray Tracing features. There are some shadow-y scenes in which RTGI and RTAO would benefit the game. But that’s just a nitpick. As I said, Stellar Blade will run smoothly on a wide range of PCs. And that’s probably what most PC gamers will appreciate.

Finally, you can go ahead and try these mods for Stellar Blade.

Enjoy!

Stellar Blade - 4K & 8K Benchmarks - Max Settings - NVIDIA RTX 5090

58 thoughts on “Stellar Blade Benchmarks & PC Performance Analysis”

  1. Leave it to John to ask about RT in an Unreal Engine 4 game.

    Is it possible to implement? Yes, with a custom library, but it does not have native support like Unreal Engine 5.

    Most developers won’t go the extra mile, especially since RT consumes so many cycles, taking resources away from other processes and significantly impacting optimization.

    1. As I said, it's a nitpick and nothing more. Since it runs so good, it would be a neat treat for those with high-end GPUs.

      1. Indeed, there are other UE4 games that had ray tracing, so it definitely is something that could be implemented. However, it might be difficult to do so if they made all the assets without RT in mind so I think they can be forgiven to skip RT options.

      1. Why would you choose to embarrass yourself?

        Unreal Engine 4 relied entirely on DXR and was a hybrid approach that incorporated rasterization, yes in past sentence.

        Epic deprecated RT experimental features in UE4 years ago due to them being not only unoptimized but also unstable, something you would know if you had worked with Unreal Engine.

        So yes, developers would need to implement a custom library to get ray tracing to work properly in UE4, as I originally stated.

        This has nothing to do with full ray tracing support like Unreal Engine 5 Lumen.

        So yes, I actually know what I am talking about.

      2. Why would you choose to embarrass yourself?

        Unreal Engine 4 relied entirely on DXR and was a hybrid approach that incorporated rasterization, yes in past sentence.

        Epic deprecated RT experimental features in UE4 years ago due to them being not only unoptimized but also unstable, something you would know if you had worked with Unreal Engine.

        So the developers would need to implement a custom library to get ray tracing to work properly in UE4, as I originally stated.

        So yes, I actually know what I am talking about, is the difference between people that read outdated articles and people that actually work with the tech.

        1. Again, what are you talking about?

          "It used DXR" – yes, obviously, like everything else.

          "UE4 RT features are deprecated" – yes, I know, Lumen does stuff differently. What's that to do with the fact UE4 games can still use RT?

          "Developers need custom libraries – no, they don't. They could do so, but it's not a requirement to actually do so.

          1. First off, when I said, “It used DXR” I was referring to the fact that Unreal Engine 4’s experimental ray tracing support was built atop DXR. However, claiming that DXR is used everywhere in ray tracing clearly shows a lack of technical knowledge, as DXR is an exclusive RT API for DirectX 12.

            Meaning yes, Unreal Engine 4 Experimental RT was exclusive to DirectX 12.

            Additionally, you don’t seem to understand what experimental, unstable, and deprecated mean. Again, developers would need a custom library to properly implement ray tracing in Unreal Engine 4.

            Since it’s clear this is beyond your scope and you lack the technical knowledge for a sound argument, I’m done wasting my time with this nonsense.

            But you free to keep entertaining me, and tell me how you running DXR on Vulkan, Metal and GNM, since you claim is used on everything, I won't reply but I'll have a good laugh.

          2. We started overall from "Leave it to John to ask about RT in an Unreal Engine 4 game.", which is such a stupid statement, given the multitudes of UE4 games that exist with RT.

            But you're so stuck up within yourself you go preaching online about being smart, roflmao.

    2. Indeed. F**k gaytracing. Waste of computational power that could be better spent in other ways that will actually make the game look better instead of killing the FPS just to proclaim your game comes with a shi*ty gimmick.

  2. I stopped buying AAA games at full price for decades. But, I am tempted to buy this game at launch to show support for the publishers and devs for this DEI free and optimized game like this. This is the way PC port should be. No unnecessary censorship, optimized for lower end hardware, no stutter. Shame we get ports like this once a year at this point

    1. >No unnecessary censorship

      One of the very first things they did was censor one of the graffiti walls that said "HARD R" because a few monkey cried about muh n word. Got revised to "CRIME R" which IMO makes it even funnier, but it was unnecessary censorship nonetheless.

      They also apparently changed up a few outfits to be less revealing but reverted the changes after enough people called them out.

      PC version also comes with denuvo.

    2. It has DENUVO, so it would be rent, not buy!
      Buuuut, if you don't care about that, go ahead, it's a good game. Nothing new, but good nonetheless!

      1. I agree that Denuvo is trash. But on the other hand if they have no DRM the game would be cracked day one and it will harm sales. Unless its a goated game like BG3 (bit woke but great game otherwise). Besides the game is fully optimized to run on 4 core cpu's. There should be no performance penalty.

        I bet they will remove Denuvo after a year or two anyways

        1. >But on the other hand if they have no DRM the game would be cracked day one and it will harm sales.

          God, this tired and idiotic argument again. How many times are we going to have to debunk this garbage? If someone doesn't want to buy something, they simply will not buy it, DRM or no DRM. Steam is DRM anyway, so denuvo is just unnecessary at this point. And take a look at the best selling PC games of all time, you'll see that some of them were DRM free from day 1. Also, used game sales hurt the profits of game studios as well yet nobody ever kvetches about that.

          1. i have to agree i have been tracking this since 2014 that denuvo came out. Good games sell despite having or not having denuvo, bad games dont despite having or not having denuvo. This is why so many bad games that flopped ended up removing denuvo, whats the point of whaving it if the game doesnt sell, since its a subscription service? It just wastes money. The issue is that there are times that denuvo broke the game or the game stopped working and devs had to take it out. So denuvo is bad.

          2. Typical denuvo trend these days since there's no more scene cracks anymore. They're just following tradition. On the other hand, Capcom are going full nuts. Adding it to 23y old game. It still sold poorly lol

          3. you talking about onimusha 2? hell you can play the switch version on switch emulator or play the better version on ps2 that has no censorship.

          4. One of the few times I write a comment that actually has paragraphs because the person I'm responding to is worth the extra effort and it gets detected as spam. Go figure. Hello, John?

          5. You, nor anyone else, have never debunked it and your comment makes you look ignorant.

            It is pure idiocy to claim that just because someone, in your case you claim literally exact 100%, of all pirates who pirate content would never ever buy it to begin with if they were willing to pirate it when the option exists had it not existed. That doesn't even make sense from the most mundane logical perspective. Would you pay for everything when you can get it free, even if you like it? No, countless people would not so that automatically debunks your entire absurd argument.

            First of all, one of the leading reasons DRM protects the initial months is because many pirates would actually, in fact, buy a game to get in on the social hype. The ability to talk about a popular game or show in person, on social media, and understand the hype around a topic is extremely powerful such that developers for movies and games spend tens of, sometimes even hundreds, of millions on marketing for a product leading up to before release. This is also why the first week or two often encompasses 40% to sometimes 90% of an entire game's lifetime sales thanks to the hype around launch.

            In fact, you don't even have a single basis to support your very silly claim.

            Those best selling games you mention have extreme piracy rates, some of the highest in history in fact. If you are claiming having piracy rates of dozens of millions did not lead to any type of significant loss of potential profit, or even any loss at all, you actually need to take some related classes and educate yourself on economics and topics like target audience, consumption behavior, and more. Marketing is far more complex than you realize.

            In fact, those particularly commonly cited "best selling games" are games that featured some of the greatest pre-launch hype seen for any video game, ever, such as Witcher 3 and Cyberpunk 2077 that helped build up their sales potential due to public interest. You also ignore that many of those highly successful ones were released on like 5-6 platforms over the life of those games, too. This is unusual and would obviously support the potential to eek out as much of the sales potential as possible.

            Used game sales are a topic that has been complained about for years as have third party key sites. It is less common a topic now as companies like GameStop fade out and digital increasingly takes over and has outpaced physical sales in recent years.

          6. Oh great, another graduate from the school of bootlicking has come in to say his two cents. You already started off on a completely invalid point, which is:

            "Would you pay for everything when you can get it free, even if you like it? No, countless people would not so that automatically debunks your entire absurd argument."

            You can already see evidence of contrary on youtube alone. Think of how many content creators put up their videos completely free of charge yet they make tons of money off their patreon. Because guess what? If people like something or someone, they will support it with money.

            I'm not going to bother responding to the rest when you already got this so hilariously wrong. Go brown-nose billion dollar corporations somewhere else, loser.

            P.S. Fvck marketing, they can all burn in the darkest pits of hell.

          7. >In fact, you don't even have a single basis to support your very silly claim.

            And yet I don't see you providing any evidence to support your drivel.

  3. Cute protagonist, beautiful characters, tight gameplay, cool soundtrack, wonderful optimization. Is that too much to ask for us PC players? And see it's UE4 that is criticized for stutters on PC. Only days gone and this game runs so well. Why? Because devs wanted these games to play well and they worked for it. I'm convinced now if any game runs poorly or has issues, it's the devs that don't want it to be, that's why.

      1. Yup! It was present since UE3. Only got worse over time. Oh and watch the new Death stranding 2 gameplay, the graphics are insane, crisp clear with decima engine. It blows UE5 out of water makes it run for money. I'm sure it will be very well optimized as well.

    1. If a game runs poorly it's usually a sign that it was made by a bunch of DEI hires that barely know what they're doing. That or more money was sunk into consultancy groups and marketing than was spent on the development budget.

      1. The present era couldn't handle beautiful women in mainstream multimedia. Imagine what nuclear explosion would it be if a stellar female beauty would've glazed a videogame in her glorious birthday suit.

  4. Congrats to all the virgins who like to fck little kids aka tr4nnies, f4ggots and weebs.
    And females

      1. Are you kidding me dude? This guy literally looks like a p3dophile and an anime enjoyer
        He also looks Jewish, coincidence? I think not

    1. The "Turd Buglar" speaks! Well barely… His nasty little way of trying to being cool…

      Easy chimp. I'm sure your family (if ya have one) can't stand you.

      Seriously fartknocker, what is it like waking up with no one who likes or cares about you?

      1. It's hard bro, I feel so lonely, is your mother available?
        probably not because you said you were gay and lgbts are social outcasts, and your parents are usually the first people who cut ties with you.

      1. Problem with that is that pedophiles like you find these attractive consistently:
        https://uploads.disquscdn.com/images/754687ae76bbcba14b5b72ac03d86f2c8de28d5d4046f0800bfeb7a7c88f805c.png

        Just look at tr4nnies and how they make themselves to look (exactly like the protagonist of this game).
        This is true while the real beauty is called d*ke by people like you:
        https://uploads.disquscdn.com/images/fdb8c8bd1143dac0a12262cd9623411423b6478c0e3bebd34cf2e753234ec30c.jpg
        You definitely have a mental illness, it's probably something to do with the fact that you have never interacted with real women. P3dophiles also find girls that look like the protagonist of this game and those ugly plastic abominations attractive as well.
        You and people like you are definitely mentally ill and I'm willing to bet my life on this statement.
        She looks wholesome? I'm pretty sure I heard that from a p3dophile that Chris Hansen interviewed 100 percent. Someone find a link to that interview please.

          1. Since when it pure Aryan beauty considered mid?
            Alright, let me see your 8, 9 or 10s

    1. Is that what photo mode is these days? Never used this "feature", social media poison everything don't it?

  5. This game has the worst CPU bottlenecks I've ever seen outside of RTS games like Civilization, which your own review mentions but for some reason doesn't carry the seriousness of the issue past 1080p as it occurs at literally every resolution / configuration after a certain tier of GPU is involved.

    Honestly, kind of amazed all the praise I see around for the game's supposed optimization when it is, by far, one of the most poorly optimized games in the past decade at the higher end of the spectrum. Yes, it scales well at the low end but quickly caps out due to the CPU wall so most players with mid to high end configurations have extremely similar performance due to this unbreachable wall. It is still good performance, regardless, but seeing GPU with a fraction of your total possible usage all the time and going "this is normal" is nuts.

  6. What order are these GPU graphs in?

    It's not GPU age chronological order
    It's not brand order
    It's not GPU average performance order
    It's not GPU 1% low order
    It's not MSRP order

    What is it?

Leave a Reply

Your email address will not be published. Required fields are marked *