Devolver Digital has just released the latest part in the Serious Sam series, Serious Sam 4, on Steam. Powered by the Serious Engine, Serious Sam 4 is one of the few games that supports both the DirectX 12 and Vulkan APIs. As such, it’s time now to benchmark it and see how it performs on the PC platform.
For this PC Performance Analysis, we used an Intel i9 9900K with 16GB of DDR4 at 3600Mhz, AMD’s Radeon RX580 and RX Vega 64, NVIDIA’s RTX 2080Ti, GTX980Ti and GTX690. We also used Windows 10 64-bit, the GeForce driver 456.38 and the Radeon Software Adrenalin 2020 Edition 20.9.1 drivers. NVIDIA has not added any SLI profile for this game, meaning that our GTX690 performed similarly to a single GTX680.
Croteam has added A LOT of graphics settings to tweak. And when I say a lot, I mean… A LOT. There is an option for enabling MSAA, resolution rendering options, Field of View slider, as well as individual settings for CPU and GPU. Unfortunately, the game does not feature any window to showcase in action what these effects do, though it does have a description for each one of them.
While Serious Sam 4 does not feature any benchmark tool, the game allows you to create your own benchmark. For both our CPU and GPU tests, we’ve benchmarked the opening level. This level features a lot of enemies on screen, and was among the most demanding scenarios we’ve experienced. We also did not experience any performance differences between Vulkan and DirectX 12. Thus, we’ve decided to stick with DX12 for our benchmarks.
In order to find out how the game scales on multiple CPU threads, we simulated a dual-core, a quad-core and a hexa-core CPU. And, when we initially tested the game, we were really disappointed by its CPU scaling. Thankfully, Croteam has issued a massive 37GB day-1 update that significantly improves CPU performance. However, you will need at least a quad-core CPU in order to play the game with 60fps. Our simulated dual-core system was simply unable to come close to a 60fps experience at 1280×720 on Ultra settings. On the other hand, our other systems had no trouble at all running the game. What’s also interesting here is the performance gains we witnessed when we enabled Hyper Threading, even on our hexa-core system. If your CPU supports Hyper Threading, we highly recommend enabling it for this particular game.
Thanks to these latest CPU improvements, we can safely say that Serious Sam 4 is a GPU-bound game. And from the looks of it, you will need a really high-end GPU in order to play it on Ultra settings, even at 1080p. At 1080p/Ultra, the only GPU that was able to offer a smooth gaming experience was the RTX2080Ti. While Croteam recommends an AMD Radeon RX Vega 64, that particular GPU was able to push a minimum of 41fps and an average of 52fps at 1080p/Ultra. So yeah, not exactly the performance I’d expect from a “recommended” GPU (unless of course the game’s PC recommended requirements are for gaming with 30fps). Not only that, but Serious Sam 4 is another game in which the Vega 64 performs poorly.
The NVIDIA GeForce RTX2080Ti was also able to provide a constant 60fps experience at 2560×1440. As for 4K/Ultra, this particular GPU was able to push a minimum of 44fps and an average of 65fps. Since there are drops below 50fps, we suggest sticking with 1440p or using a custom 4K resolution.
Graphics-wise, Serious Sam 4 is a mixed bag. The in-game graphics look – for the most part – great, with some nice textures and effects here and there. You can dismember all the enemies, and there is limited environmental destructibility. However, the cut-scenes look awful and perform poorly. It feels like some cut-scenes have wrong shading effects, something that makes them feel worse than even Serious Sam 3. Not only that, but the indoor environments look mediocre. There is also a lot of pop-in, even on Ultra settings. Thankfully, you can minimize it, though you will still experience some distant pop-ins. We also experienced some flickering and visual artifacts. And, to be honest, for what is being displayed here, the game should be running better.
In conclusion, Serious Sam 4 performs pretty good on NVIDIA’s Turing architecture. The game also appears to be scaling great on multi-core CPUs. However, performance on every other graphics card, whether it’s AMD’s or NVIDIA’s older architectures, is underwhelming. Furthermore, the game’s graphics may disappoint some PC gamers, especially since they are not that much better than Serious Sam 3; a game that came out nine whole years ago. To be honest, this game is a mess so we advice you to wait until – and if – Croteam fixes it.
John is the founder and Editor in Chief at DSOGaming. He is a PC gaming fan and highly supports the modding and indie communities. Before creating DSOGaming, John worked on numerous gaming websites. While he is a die-hard PC gamer, his gaming roots can be found on consoles. John loved – and still does – the 16-bit consoles, and considers SNES to be one of the best consoles. Still, the PC platform won him over consoles. That was mainly due to 3DFX and its iconic dedicated 3D accelerator graphics card, Voodoo 2. John has also written a higher degree thesis on the “The Evolution of PC graphics cards.”
Contact: Email
Serious Sam 3.5
From what I’m reading the graphics need some serious optimization. A 2080 Ti being necessary for ultra on 1080p isn’t going to be alright with the vast majority of gamers that can’t afford a 2080 TI. Cut-scenes look awful, indoor environments look mediocre, a lot of pop-in, flickering and visual artifacts.
No, this is not acceptable.
RTX2080Ti was CPU bottlenecked at 1080p (notice the same framerates at both 1080p and 1440p), so the game can run with 60fps even on older Turing models. Curious to see whether the performance has been improved on the GTX980Ti and AMD’s GPUs after the latest update.
No other game as that many enemies on screen. Its the amount of bad guys. Most games have far far less. Do you want them to have like 4 bad guys like most games so then you can gawk at how smooth the game runs.
Man I was really hoping this game would be more like SS1 FE, SS1SE and SS2, but it really is just SS3 with a new map. Like SS3 the framerate is bad, the maps are uninspired and look like crap, the weapons are just reskins, you have to reload weapons, there is aiming down sights, lighting is mostly crap, there are a lot of bugs, the melee combat is useless, the npc’s were made in 2006 and it’s overall just less fun then the original games.
Dual double-barreled shotguns, dual miniguns? Should be senseless, silly fun, I think I won’t be disappointed.
I just hope it wont end like Postal 3….
Its getting bad reviews from all kinds of people that loved the past games.
No new interesting enemies, bad lvl design, big battles are actually rare, tons of boring story you didn’t buy the game for etc…
I was so looking forward for this, especially because the developers are from my country and I know someone that works there.
I’ve read several reviews, and they are actually mixed. I never expected this game to get rave reviews.
I also checked some gameplay footage, and it’s about what I thought it would be.
Reportedly the game is around 10-12 hours long, so it will be over before I get bored. I’m pretty sure it will be fun for me, but I’m gonna wait for a sale.
You must not own the game then. Play it on 1080P, ultra cpu, medium GPU and it will be pretty and smooth as silk if not something wrong with your system.
Sad to see such a dated looking game needs so much power.what a joke.
This is a “old school” fast shooter game and that was not their aim, also I wasnt expecting a AAA level of graphics but sure, this needs to be better (look at those reflections on water), texture poppins are not acceptable etc…. it should at least look decent and clean, not like a “mess”.
Every developers aim must be an optimized game. compare doom to this crap and see my point. the game requirements are ridiculous.
There are thousands of enemies on screen. You are a joke if you don’t understand the cpu requirements. These guys are daring to say f*k it we are going to make a game good and the cpu guys need to make something to keep up. Like crysis for the cpu.
Most people don’t understand that. They don’t even play the game they just benchmark them and cry.
For me it plays smoother than Borderlands 3 and that plays over 100FPS on my 1080 GTX. Borderlands 3 does not have 1/10 the onscreen enemies either.
So much for AMD’s fine wine
“In conclusion, Serious Sam 4 performs great on NVIDIA’s hardware”
“At 1080p/Ultra, the only GPU that was able to offer a smooth gaming experience was the RTX2080Ti”
lol.. just lol.
rtx2080ti here and it disagrees
Those system requirements were a cover up for bad optimization after all… who knew.
Croteam released another 1GB Update, so I’ll be re-testing the GPUs tomorrow (and I’ll update the graphs in case there are any improvements).
Edit: There are no performance improvements at all on both AMD’s and NVIDIA’s hardware.
Thank you John.
@JohnDio:disqus isnt this one of the strangest CPU scallings we have seen….. this game really likes threads instead of cores. Mostly its the other way around or some specific games just generally dont like threads (doesnt matter if you have many or less threads) and there are other games which are good optimized but those games like both cores and threads.
I wondered why my 3770k + gtx 1080 didn’t like 1440p ultra w/ 4k mpixel whatever that setting is. ????
what the…. 1080p ultra max all with smooth 60 fps, is only for rtx 2080 ti 11gb …… T_T
This franchise should have died by now, as it’s not shown any growth whatsoever.
Thats strange I have the game day 1 and play it on an older intel hex core + GTX 1080 and its no lag smooth as silk. The cpu at ultra and gpu at medium. On a older 970 gtx it lags a bit unless you take off FSAA. The graphics are way better than Serious Sam 3. Its not a mess at all.
Nice job with all the detailed info, John! ? A lot of settings to tweak means (hopefully) a lot of options to tweak the performance of the game. How does the performance and visuals look compared to lowest and visually optimized settings? I’m curious about diminishing returns.
As is the case with most of the games, there are points where there is absolutely no reason to increase the settings further just because the developers added that option in the menu for future hardware and benchmarking geeks. Very few reviewers remember about this and mostly care about maximum numbers. It would be interesting to see how all the settings are optimized in any game (not just the maximums).