Earlier today, Stu reported Denuvo Anti-Tamper DRM protection had been patched out of Hitman 2. Naturally, my first port-of-call was to re-download Hitman 2 and see whether the removal of Denuvo had any impact on Hitman 2's frame rates.
As always with Denuvo benchmarks though, we’re all sat around trying to prove the unprovable. It would take an immense out of luck (or dedication) to have benchmarked Hitman 2 yesterday before the removal of Denuvo. As it is, we’re faced with two separate benchmark results performed on the same hardware but separated by three months, several patches, and multiple driver updates.
However, even with new graphics card drivers, and the removal of Denuvo, it’s hard to argue a case for improved frame rates in Hitman 2, as you can see in the Hitman 2 benchmarks with and without Denuvo DRM below.
Hitman 2 Denuvo Benchmarks
These Hitman 2 benchmark results have been gathered from two separate tests, both using the exact same hardware - a GeForce GTX 1060 6GB, Intel Core i7-5820K CPU, and 16GB DDR4 memory.
The benchmark tests for Hitman 2 with Denuvo were performed back in November, while the Hitman 2 without Denuvo scores were benchmarks taken today.
Both tests were the same run on the same Miami level in Hitman 2.
Hitman 2 1080p Denuvo Benchmarks
Hitman 2 1440p Denuvo Benchmarks
Hitman 2 4K Denuvo Benchmarks
| 1080p With Denuvo | 1080p Without Denuvo | 1440p Denuvo | 1440p Without Denuvo | 4K Denuvo | 4K Without Denuvo | |
| Low | 89 | 78 | 66 | 68 | 34 | 39 |
| Medium | 80 | 75 | 57 | 57 | 30 | 31 |
| High | 72 | 73 | 47 | 51 | 27 | 26 |
| Ultra | 68 | 65 | 44 | 49 | 26 | 25 |
If can draw any concrete conclusions out of these figures then you're a better statistician than I. The area where we'd expect to see Denuvo have an impact is at lower resolutions and with Low visuals in Hitman 2, but we actually end up with the opposite. Whether that's due to changes in population density or something on Io Interactive's end we just don't know, but at the settings where we would expect Denuvo to cause an FPS hit if it causes one at all, the opposite happened.
Aside from this minor anomaly, there's honestly so little at play here that the performance difference in these benchmark runs comes entirely down to margins of error. Hitman 2 runs pretty much the same now without Denuvo as it did three months ago with Denuvo.
The results may bear more fruit while using a much weaker CPU but any program using CPU cycles will by definition impact a weaker CPU more than a strong one. Hitman 2 isn't exactly a game that goes easy on your processor either, so those with low-end CPUs will probably find themselves struggling for stable frame rates even with Denuvo removed.
Well, that's the results. Again, there's no conclusive answer and I don't think we're going to get one until we're actually forewarned about Denuvo being removed from a game.
If any of you folks happen to have benchmarked Hitman 2 very recently then we’d also to love to hear your experiences with it now Denuvo has been removed!
What are your thoughts on the results? Is this the sort of FPS impact you were expecting? Let us know!
Login or Register to join the debate
PC Specs
on 4k the difference is small.
i donot know why may be it does not work on higher resolutions.
PC Specs
Speaking of denuvo - Resident Evil 2 was cracked few hours ago.
"Insane protections need insane people defeating it"
PC Specs
reading nfos are we?
PC Specs
Always-)
PC Specs
Didn't RE:2 removed denuvo?
PC Specs
nope, it is cracked
PC Specs
Okay. I guess it was on other game that was removed or I'm mistaken.
PC Specs
yeah that was re7 or idk the one before this
PC Specs
mm123456 It was Resident Evil 7. Resident Evil 2 Remake still has Denuvo implemented
PC Specs
yea this was it. Ty.
PC Specs
As an indie developer guys i can tell you that no matter what you add to the engine from a small script to a huge mesh ( as in many verticles) does affect overall performance
PC Specs
It's pretty obvious this testing methodology is flawed. Running around in an area isn't a reliable way to make a chart because you can easily take the data of the best Denuvo run and compare it to the worse non Denuvo run.
PC Specs
Interesting results
PC Specs
I never really cared if it effected performance or not. I hate the fact that it's treating paying customers like thieves. People that pirate the game don't have to deal with it, only paying customers. DNA swabs and house searches can be unobtrusive as well and yet, here in the states, they can't be conducted with out good cause.
PC Specs
Speaking of DNA swabs... those DNA ancestry companies (23 and Me, Ancestry, etc) are selling (stupid) people's DNA to pharmaceitical companies for biiiig money, which is why those DNA ads started popping up all over the place within the last year or two. Good business. Screw the people :D
PC Specs
What about the 1% lows average means nothing. 44 fps could mean a stuttering mess or a smooth experience. Also notice how long the graphs are so a frame difference would looks bigger.
PC Specs
My thought also. If the "anti-tamper" is looking at files, encrypting, decrypting, calling home, etc - how would that impact the performance in other ways than just the fps counter..? Nobody really talks about that and I do wonder if there's an impact.
PC Specs
I had the same issue with Norton virus scanner my averages weren't much lower but my 1% and 0.1% lows were affected in a significant way that microstutter was noticeable. Uninstalled it and learned to live with the sufficient windows def.
PC Specs
[Conspiracy theory]
What if Devs and Publishers themselves are capping the performance of games with Denuvo removed, or even if they are not actually removing Denuvo ?!
Doing this, when tests of this type are made, the performance difference will be minimal between one version and another... Decreasing Rage against them when they are launching new games using this DRM.
Think about it
PC Specs
Or not... I did not take my remedies today
PC Specs
I feel as though this only confirms my suspicions, that Denuvo doesn't actually hurt performance.
The real issue here then is the ability to play the game at all should Denuvo's servers kick the bucket in the future.
PC Specs
Conclusion: Denuvo's performance is the best at 1080p.
PC Specs
LOL
PC Specs
Awesome!
PC Specs
Performance wasn't ever the issue for me for any denuvo game but its the server denuvo needs that bothers me the most, i look at it like gfwl, if shutdown on there end we have to either beg for it to be removed (which might not happen for all) or go down the crack route
Saying that though i do feel as if this game now loads faster on the loading screens
PC Specs
Suppose the only real way to measure this is to do it on the same driver and the same game version...legit (with denuvo) and pirated (without)...
Otherwise we just don't know what changed in the drivers, Windows updates, your bench system or the game builds.
PC Specs
You're not wrong. It felt pointless the entire time I was doing it lol, especially when there wasn't some big change
PC Specs
I guess at this point I'd be more interested in more subtle ways Denuvo could be impacting gameplay. Things like potential network load while calling, game load times (encryption/decryption), if any additional stuttering is caused by any Denuvo "anti-tamper" checks, quality of life changes (modding and tweaking capability, installing/re-installing the game multiple times for whatever reason), etc., rather than raw framerate digits. We've seen a few times so far that these numbers are more or less within margin of error.
Regardless, there's no shadow of a doubt that not having Denuvo is still the ideal scenario for all sorts of reasons, especially historical game preservation.
PC Specs
True, but you won't see that much of an increase from drivers, especially for a game that came out 3months ago already. Usually new drivers improve stability and/or game day drivers, it won't be a driver in 3months that gives Hitman 2 better fps. AMD does seem to do way better with drivers and improvements then Nvidia does. But i do agree should be done the same day, all we can see here is some of the results are even within a margin of error. Denuvo has pretty much very little impact.
PC Specs
Sometimes newer drivers include bugfixes for older games, for example if crashes or flickering occur after the WHQL is released for that game. In that case the fix may improve or decrease the performance eeeeever so slightly, because the code obviously is doing something different, even if for that one set of draw calls.
From my understanding Nvidia allegedly already gives you the best performance out of the box with WHQL drivers, whereas AMD is notoriously a bit wonky sometimes. So that then begs a question whether this "fine wine" that AMD is praised for just means they're optimizing and fixing their past mistakes? Or is it that they invent some new optimizations and Nvidia does not..?
PC Specs
I love how people still think AMD drivers are wonky lol. I've heard far more horror stories with nvidia in the past year then AMD drivers. I use both on builds i do and i usually have more people contact me with nvidia driver issues then AMD ones. There is no AMD wonky drivers anymore, hasn't been the case in years.
PC Specs
I see the newer Adrenalin drivers being wonky with my Pulse RX 470 (I'm now stuck with Pre-19.1.1 driver for proper usability). Also just head to AMD forums and see people complain about the "new and improved" fan-curve being weird (revving up and slowing down constantly, killing fan-stop function), Relive recording weird colours, P-states being reset in Wattman, Performance Metrics sometimes not showing in full-screen games (but always work in borderless window), Adrenalin recommends driver update that is older than the one already installed (and the driver update->reset->update to older driver->loop continues), etc
PC Specs
That's interesting, Ive not run into any of those issues with my v64 or my r9 390 on my other rig. Ill have to give it a test, i dont use radeon to monitor i use ab as i always have so ill give it a go. I wonder how many people actually tried the new 19.2.1 that fixes all the fan issues. From what ive seen on oc.net a lot of people have it working just fine.
PC Specs
For me the new driver effectively disables any fan-speed control even in Afterburner, while the older drivers had no such problem. I googled to see if anyone got the same issue and sure enough the reddit threads complain about the same thing popped up first in search result. Personally this has been the worst period of AMD driver for me, previous Adrenalin and Crimson versions worked great.
PC Specs
KID, try the new 19.2.1 if you haven't yet, came out a couple days ago. I haven't had any issues on my V64 but i have heard of people having fan issues on the rx series, then I've had people tell me their fan control works fine. I'm assuming it's clashing with afterburner but not sure, I've had crimson and adrenalin drivers do that before as well.
PC Specs
I'll try it, thanks. Hopefully it works.
I only tried AfterBurner (with Wattman uninstalled) after fan control in Adrenalin keeps revving the fans up and down. Normally I found the AMD driver to be sufficient without AfterBurner, seemed like the driver couldn't control the fans regardless of software used.
PC Specs
The pirated version isn't the same as magically removing denuvo. The protection is still there, the exe was still like 250mb it was working but bypassed
PC Specs
"Working, but bypassed" sounds to me like "the door was locked, but I went through". Denuvo either works or doesn't.
Well, I suppose if we're spoofing their server check locally somehow then there's no positive change, but if the code is there and is simply bypassed as you say (read: code is left unused) then Denuvo is, essentially, gone.
PC Specs
they say that denuvo is still running it just gets "tricked" by the crack
PC Specs
exactly, the latest cracks come with untouched game exes, the protection is tricked to think that everything is fine
PC Specs
I don't think cracked versions remove Denuvo, they just fool it into thinking it's communicating with home servers. As far as I understand, Denuvo is still running on a cracked version just without access to the internet.
Nevertheless I would like to see a side by side comparison of a game with, without and cracked one. I still doubt that'd show any significant or noticeable difference and I don't think anyone would really have fun doing it.
PC Specs
If crapnuvo is still running but pinging a spoofed "home server" then yeah, probably no difference. And no modding, because it would crap itself.