Hitman 2 Benchmark Performance Analysis

[W1zzard]

Hitman 2 is the latest installment in the epic contract-killer series. IO Interactive has expanded on their winning concepts from the previous title with larger, more detailed maps giving you even more unique ways to take out your target. We tested the game's performance using all the latest graphics cards, including GeForce 20 RTX and RX 590.

Show full review

 

[kastriot]

So AMD owners are left in the mud because no DX12 or Vulkan is available and performance sucks in DX11 especially for RX series so this makes this game Nvidia exclusive.

 

[M2B]

Based on my own experience most of the time framerate is significantly higher than what W1zzard is showing us here but I guess that's a good thing because this seems like a true stress test.
If you have a G-Sync monitor RTX 2070 should be sufficient at 4K with a good overclock and sacrificing the level of detail a little bit.

 

[Kamgusta]

So AMD owners are left in the mud because no DX12 or Vulkan is available and performance sucks in DX11 especially for RX series so this makes this game Nvidia exclusive.

Nobody asked AMD to produce VGAs that poorly perform in DX11.

 

[bug]

Dropping DX12 from a title/engine that previously supported it is a big ouch to DX12's adoption.
Also, 1060 faster than the newly released 590. That's an extra $50 well spent

 

[Sokol94]

Interesting. Because Hardware Unboxed got quite different results from you guys:

Basically 570 matches 1060 3gb, 580 matches 1060 6gb and then the cards between gtx 1070 till rtx 2080 are all within few fps difference in both min and avg fps.

I don't own the game so cannot confirm but some reddit users mentioned the game now has hairworks option. Was it enabled? If so that would explain the gimped performance on amd cards.

 

[W1zzard]

the game now has hairworks option. Was it enabled?

I haven't seen any hairworks option. Settings screenshot is there for you to check out.

Any chance the video used the warez version and not the regular public version? Did they use the integrated benchmark? I also used the latest drivers

 

[Sokol94]

I haven't seen any hairworks option. Settings screenshot is there for you to check out.

Any chance the video used the warez version and not the regular public version? Did they use the integrated benchmark? I also used the latest drivers

I cannot comment on the version of the game as they did not disclose. They took average results from three runs lasting 60 seconds. The bench included (quoted from video):
"for testing I'm using the Miami mission titled the finish line and my save game location starts us off at the npc heavy grandstand for the tests I walk the entire length of the grandstand then exit to the right and then take another right and then the test finish is basically there".

So it could be the case of some locations in the game being more taxing than the other?

 

[bug]

I cannot comment on the version of the game as they did not disclose. They took average results from three runs lasting 60 seconds. The bench included (quoted from video):
"for testing I'm using the Miami mission titled the finish line and my save game location starts us off at the npc heavy grandstand for the tests I walk the entire length of the grandstand then exit to the right and then take another right and then the test finish is basically there".

So it could be the case of some locations in the game being more taxing than the other?

Why on earth are you talking a custom benchmark and try to compare it to results elsewhere? What's the point of that?

 

[W1zzard]

So it could be the case of some locations in the game being more taxing than the other?

Yup, I just checked through the maps quickly and there's huge differences

 

[Capitan Harlock]

I run the game all right with my Fury X . I have SSAO off , shadow standard and all on extra with vsync on.
Seems like Freesync is not working right for now so i turned that off .

 

[TheGuruStud]

So AMD owners are left in the mud because no DX12 or Vulkan is available and performance sucks in DX11 especially for RX series so this makes this game Nvidia exclusive.

More like lazy devs or bribed. Take your pick. Perf is garbage, regardless.

 

[rtwjunkie]

More like lazy devs or bribed. Take your pick. Perf is garbage, regardless.

Did you read the review? It doesn’t sound like performance is garbage. And W1z makes a special mention that AMD cards did better as resolutions got higher.

 

[TheGuruStud]

Did you read the review? It doesn’t sound like performance is garbage. And W1z makes a special mention that AMD cards did better as resolutions got higher.

Did we look at the same numbers. It's basically unplayable unless you blew a wad of dough on turding. That's unacceptable. EVEN 1080P IS UNPLAYABLE! Same thing once you step to 1440 with that range of cards.
You will need 1080ti perf to maintain stable framerate at 1440 (which probably still stutters in crowds lol). That's absurd. The first one was demanding enough, but at least I can play at 1440 on 980ti. RIP. I won't even be pirating it.

 

[rtwjunkie]

More like lazy devs or bribed. Take your pick. Perf is garbage, regardless.

Oh, so 63.8fps on 1440p is unplayable. Got it.

 

[TheGuruStud]

Oh, so 63.8fps on 1440p is unplayable. Got it.

You do realize 60 has to be MAINTAINED. This is AVERAGE, meaning there's A LOT of frames under 60. Unplayable. In this game that means lots of dips.

 

[bug]

More like lazy devs or bribed. Take your pick. Perf is garbage, regardless.

I have said from the very beginning both DX12 and Vulkan being lower level, makes them akin to C. You can do a lot of things with it, but most of the time it just won't be worth the effort. But no, dev must be lazy or bribed.

Did you read the review? It doesn’t sound like performance is garbage. And W1z makes a special mention that AMD cards did better as resolutions got higher.

Well, at 4k Vega 64 finally got a nice lead on the 1070 and 590 finally overtook the 1060. That's fine, but at 4k none of those card were in playable territory anyway.
I've asked before when doing aggregate performance graphs, to add a second set of graphs including only titles were the tested card were actually playable.

 

[TheGuruStud]

I have said from the very beginning both DX12 and Vulkan being lower level, makes them akin to C. You can do a lot of things with it, but most of the time it just won't be worth the effort. But no, dev must be lazy or bribed.

Well, at 4k Vega 64 finally got a nice lead on the 1070 and 590 finally overtook the 1060. That's fine, but at 4k none of those card were in playable territory anyway.
I've asked before when doing aggregate performance graphs, to add a second set of graphs including only titles were the tested card were actually playable.

On a game performing this shitty, it sure is worth it unless you just don't care, b/c you're lazy and failing (or a financial interest has no interest in dx12). It IS one or the other.

 

[rtwjunkie]

You do realize 60 has to be MAINTAINED. This is AVERAGE, meaning there's A LOT of frames under 60. Unplayable. In this game that means lots of dips.

Do tell? Gosh, I had NO idea.

From the Conclusion page:
“ a GTX 1070 Ti / 1080 / 2070 will give you a great 60 FPS gaming experience at 1440p”.

 

[bug]

On a game performing this shitty, it sure is worth it unless you just don't care, b/c you're lazy and failing (or a financial interest has no interest in dx12). It IS one or the other.

Yes, because so far we've had a ton of games that have performed so much better in DX12.

For the time being, one simply cannot target just DX12, because that's Win10 only. And if you're just going to slap DX12 on top of DX11, there goes all the (good) reasons for going DX12. But what do I know, you seem to have all the answers already.

 

[Valent117]

why you don't do cpu scaling benchmark anymore?

 

[Aldain]

Um w1zzard your results are wayyyyy different from every other results out there

 

[renz496]

More like lazy devs or bribed. Take your pick. Perf is garbage, regardless.

so if developer are not using stuff that you want them to use they are automatically labelled as lazy or being bribe by someone else?

 

[Ferrum Master]

Hairworks for Agent 47, lul wut?

 

[newtekie1]

On a game performing this shitty, it sure is worth it unless you just don't care, b/c you're lazy and failing (or a financial interest has no interest in dx12). It IS one or the other.

The Hitman games have never performed well, but they don't need to. The games aren't fast action games, if you are playing correctly. The games are actually smooth down to 30FPS. I can't comment exactly on this one, because I don't own it yet, but I can say all the other games are smooth at 30FPS.

The crap about anything below 60FPS is unplayble is bullsh*t. Same games don't need to be at 60FPS to be smooth. Turn the FPS counter off and enjoy the game.