News Hitman 3 Benchmarks and Performance

[Admin]

We benchmarked Hitman 3 on the latest graphics cards to see how it runs, and we'll be adding more results over the coming days.

Hitman 3 Benchmarks and Performance : Read more

 

[UWguy]

Supposedly they are adding ray tracing at a later date. Please re-benchmark and let us know if he is still wearing red then. Thanks.

 

[Math Geek]

usually enjoy these benchmarks articles, but to only include cards that can't be bought makes it lose some of the fun of it

maybe a mid-range and budget card would be useful as well to see if someone who did not spend $1500+ for their pc can play the game!!

 

[TCA_ChinChin]

Game runs a lot better on Nvidia GPUs: Perfectly normal, no need to call it out. Happens all the time.
Game runs a lot better on AMD GPUs: "Agent 47 is dressed up in AMD red".

 

[benderman]

I just registered to call out this article big BS right there. Just check other outlets, the AMD cards are 3-5% behind nvidia. Just as an example, look at Computerbase for the same resolutions. Not sure what you guys did but AMD cards aren't hitting 200+ fps anywhere.

 

[MasterMadBones]

I just registered to call out this article big BS right there. Just check other outlets, the AMD cards are 3-5% behind nvidia. Just as an example, look at Computerbase for the same resolutions. Not sure what you guys did but AMD cards aren't hitting 200+ fps anywhere.

Waiting for Jarred to confirm this, but AMD released a driver update that's supposed to boost performance in Hitman 3 by up to 10%. There may be a discrepancy in the driver version used by other outlets.

Edit: it also seems that I can't read.

 

[JarredWaltonGPU]

I just registered to call out this article big BS right there. Just check other outlets, the AMD cards are 3-5% behind nvidia. Just as an example, look at Computerbase for the same resolutions. Not sure what you guys did but AMD cards aren't hitting 200+ fps anywhere.

Gee, it couldn't possibly be because I used a different CPU, could it? Not sure which test sequence was used there either. But considering it's Intel promoted, yeah, Core i9-9900K might actually beat Ryzen 9 5950X. It happens.

 

[vilor]

I just registered to call out this article big BS right there. Just check other outlets, the AMD cards are 3-5% behind nvidia. Just as an example, look at Computerbase for the same resolutions. Not sure what you guys did but AMD cards aren't hitting 200+ fps anywhere.

Guru3d also has Amd winning this game. 5950X cpu. also 6800 is virtually a tie with 3080 same as here.
Maybe computerbase article is wrong... what other articles you found out?
You said Computerbase was just an example care to through another?
Maybe you got angry for another reason and had to make an account to say this instead of checking more articles.

 

[benderman]

Guru3d also has Amd winning this game. 5950X cpu. also 6800 is virtually a tie with 3080 same as here.
Maybe computerbase article is wrong... what other articles you found out?
You said Computerbase was just an example care to through another?
Maybe you got angry for another reason and had to make an account to say this instead of checking more articles.

It rather seems to me that some of these tests try to pander to a certain audience. Guru3D has only 10-15fps advantage for AMD at 1080p, not 30 fps like in this test. It also seems to me that the performance comparison is heavily dependent on the scene. Computerbase does not use the Dubai nor the Dartmoor scenes and Nvidia comes out ahead.
There's absolute no logic a 6800 XT being 30 fps ahead of 3090.
The other tests I was referring to were videos all over Youtube. I don't see any huge gap in performance like here.

 

[JarredWaltonGPU]

It rather seems to me that some of these tests try to pander to a certain audience. Guru3D has only 10-15fps advantage for AMD at 1080p, not 30 fps like in this test. It also seems to me that the performance comparison is heavily dependent on the scene. Computerbase does not use the Dubai nor the Dartmoor scenes and Nvidia comes out ahead.
There's absolute no logic a 6800 XT being 30 fps ahead of 3090.
The other tests I was referring to were videos all over Youtube. I don't see any huge gap in performance like here.

Here's the logic:

AMD has traditionally good DX12 performance. Navi 21 has a 128MB L3 "Infinity Cache" that some games really respond well to. The built-in Dubai benchmark may not be as demanding as other areas of the game. If you play in an area that has more NPC stuff going on, you will become more CPU limited, which in turn means you're less likely to see big differences between the GPUs. You're also looking at the 1080p results, which is the best-case scenario for AMD's large cache. Most places probably didn't run at medium and ultra at 1080p/1440p/4K on all of the GPUs. The fastest GPUs will all hit CPU bottlenecks at 1080p, which means drivers and other factors come into play.

Look at GPU limited resolutions (mostly 4K in this game) and the 3090 is basically tied with the 6800 XT. There are other games where we see similar results. Assassin's Creed Valhalla for sure, Borderlands 3 at 1440p, and Forza Horizon 4 are all examples of AMD's 6900 XT leading the 3090. New drivers could improve Nvidia's standings. Game patches could do that as well. Maybe in the game itself (rather than built-in benchmark) Nvidia does better. Investigating every potential factor takes time, though.

These are early results, and they're valid for what they show. Are they universally applicable to the game as a whole? I don't know yet, but in general it isn't even a major factor. You can't buy these cards, and they're all running at well over 60 fps even at 4K and max settings -- in a relatively slower paced stealth game.

 

[King_V]

Ah, "videos all over YouTube"

Yes, CLEARLY reliable authorities on the matter. 🙄

 

[Conahl]

maybe instead of posting videos all over YouTube, he could post the links to said videos, so we can be sure we have the ones to which he refers ?

 

[namtrooper81]

Would be nice to see some 1080ti performances. Guess it is labeled "too old" now. On average I normally just look at 2080 scores to do a comparison.

 

[benderman]

Here's the logic:

AMD has traditionally good DX12 performance. Navi 21 has a 128MB L3 "Infinity Cache" that some games really respond well to. The built-in Dubai benchmark may not be as demanding as other areas of the game. If you play in an area that has more NPC stuff going on, you will become more CPU limited, which in turn means you're less likely to see big differences between the GPUs. You're also looking at the 1080p results, which is the best-case scenario for AMD's large cache. Most places probably didn't run at medium and ultra at 1080p/1440p/4K on all of the GPUs. The fastest GPUs will all hit CPU bottlenecks at 1080p, which means drivers and other factors come into play.

Look at GPU limited resolutions (mostly 4K in this game) and the 3090 is basically tied with the 6800 XT. There are other games where we see similar results. Assassin's Creed Valhalla for sure, Borderlands 3 at 1440p, and Forza Horizon 4 are all examples of AMD's 6900 XT leading the 3090. New drivers could improve Nvidia's standings. Game patches could do that as well. Maybe in the game itself (rather than built-in benchmark) Nvidia does better. Investigating every potential factor takes time, though.

These are early results, and they're valid for what they show. Are they universally applicable to the game as a whole? I don't know yet, but in general it isn't even a major factor. You can't buy these cards, and they're all running at well over 60 fps even at 4K and max settings -- in a relatively slower paced stealth game.

I looked up the test sequences from Computerbase extra for you my friend and they prove you wrong. They separated both CPU and GPU heavy sequences. Their GPU sequence has almost no NPC's at all unlike the Dubai scene which may explain the big difference if this game uses a lot of CPU cores (which then favors Ryzen of course and has nothing to do with Nvidia).
I don't know if the benchmark sequence you all using is favoring AMD or you got a quick buck for it, but you can believe what you want.
As for myself, I will rather trust the Germans at Computerbase.
Good night.

 

[TheBeastInside]

So to summarize, no discernible performance differences in the vaporware sector.
Let's definitely argue the minute meaningless differences..

 

[Makaveli]

I looked up the test sequences from Computerbase extra for you my friend and they prove you wrong. They separated both CPU and GPU heavy sequences. Their GPU sequence has almost no NPC's at all unlike the Dubai scene which may explain the big difference if this game uses a lot of CPU cores (which then favors Ryzen of course and has nothing to do with Nvidia).
I don't know if the benchmark sequence you all using is favoring AMD or you got a quick buck for it, but you can believe what you want.
As for myself, I will rather trust the Germans at Computerbase.
Good night.

I don't understand why you even bothered to create an account since you only trust "Computerbase"

You are calling out bias yet Guru numbers matches what THG saw.

 

[JarredWaltonGPU]

Would be nice to see some 1080ti performances. Guess it is labeled "too old" now. On average I normally just look at 2080 scores to do a comparison.

More like: each GPU takes time to test and I started at the top. I'll probably run a few GTX 10-series numbers (along with AMD RX Vega and 500-series) before calling it a day. But it's Saturday and people aren't super interested in Hitman 3 benchmarks (based on traffic), so I'm not going to kill myself running tests all weekend.

I don't understand why you even bothered to create an account since you only trust "Computerbase"

You are calling out bias yet Guru numbers matches what THG saw.

Some people can't see past the nose on their face. Referencing a site with different results that didn't explicitly list every aspect of how they tested (I'm pretty sure the article has been updated with additional info now) and comparing it to my results, and then suggesting mine were "paid for" while the other site's are "correct" is the height of hypocracy. And yet, it's the internet so we see that garbage all the time. To be clear: I am NOT saying Computerbase's results are wrong. They're different, but that's because they tested in a different fashion on different hardware. They're just one more data point to consider.

Anyway, I have a general rule: I'll point out flaws in reasoning once when a person tries to suggest my results are "incorrect" just because they don't like what they show. After that, it's the ignore button -- very handy for avoiding stupid trolls.

 

[Math Geek]

so true so true.

nicely said

 

[helper800]

More like: each GPU takes time to test and I started at the top. I'll probably run a few GTX 10-series numbers (along with AMD RX Vega and 500-series) before calling it a day. But it's Saturday and people aren't super interested in Hitman 3 benchmarks (based on traffic), so I'm not going to kill myself running tests all weekend.

Some people can't see past the nose on their face. Referencing a site with different results that didn't explicitly list every aspect of how they tested (I'm pretty sure the article has been updated with additional info now) and comparing it to my results, and then suggesting mine were "paid for" while the other site's are "correct" is the height of hypocracy. And yet, it's the internet so we see that garbage all the time. To be clear: I am NOT saying Computerbase's results are wrong. They're different, but that's because they tested in a different fashion on different hardware. They're just one more data point to consider.

Anyway, I have a general rule: I'll point out flaws in reasoning once when a person tries to suggest my results are "incorrect" just because they don't like what they show. After that, it's the ignore button -- very handy for avoiding stupid trolls.

I'm glad you at least try to reason with them, however, don't let this kind of thing drag you down or bother you in the slightest. This site, in my opinion, is beyond reproach as far as credibility is concerned.

 

[benderman]

More like: each GPU takes time to test and I started at the top. I'll probably run a few GTX 10-series numbers (along with AMD RX Vega and 500-series) before calling it a day. But it's Saturday and people aren't super interested in Hitman 3 benchmarks (based on traffic), so I'm not going to kill myself running tests all weekend.

Some people can't see past the nose on their face. Referencing a site with different results that didn't explicitly list every aspect of how they tested (I'm pretty sure the article has been updated with additional info now) and comparing it to my results, and then suggesting mine were "paid for" while the other site's are "correct" is the height of hypocracy. And yet, it's the internet so we see that garbage all the time. To be clear: I am NOT saying Computerbase's results are wrong. They're different, but that's because they tested in a different fashion on different hardware. They're just one more data point to consider.

Anyway, I have a general rule: I'll point out flaws in reasoning once when a person tries to suggest my results are "incorrect" just because they don't like what they show. After that, it's the ignore button -- very handy for avoiding stupid trolls.

Since you called me a stupid troll breaking news just in for you: yet another outlet has almost same numbers as Computerbase. Just look up PC Games Hardware test that came out today. Nowhere near the same 30fps differences as Tom's Hardware test. The 3090 is actually ahead in WQHD and on par with 6900 XT in 4K. At 1080p the difference between 3090 and 6900 XT is 2! fps.
I'd rather call this website and it's users trolls for posting ridiculous numbers to trick their audience without any peer-to-peer checks.

 

[Makaveli]

Since you called me a stupid troll breaking news just in for you: yet another outlet has almost same numbers as Computerbase. Just look up PC Games Hardware test that came out today. Nowhere near the same 30fps differences as Tom's Hardware test. The 3090 is actually ahead in WQHD and on par with 6900 XT in 4K. At 1080p the difference between 3090 and 6900 XT is 2! fps.
I'd rather call this website and it's users trolls for posting ridiculous numbers to trick their audience without any peer-to-peer checks.

Close the door on the way out dude.

Complaining because you didn't like NV's position in the results that is the only reason you created an account.

Fan boys so rampant these days. Anyone that has actually been doing this for awhile knows to read multiple reviews before forming an opinion.

 

[benderman]

Close the door on the way out dude.

Complaining because you didn't like NV's position in the results that is the only reason you created an account.

Fan boys so rampant these days. Anyone that has actually been doing this for awhile knows to read multiple reviews before forming an opinion.

Yeah, thanks will do. I read 2 reviews from reputable sources that confirm my point.
Good luck to you!

 

[Makaveli]

Yeah, thanks will do. I read 2 reviews from reputable sources that confirm my point.
Good luck to you!

two reviews is not enough to form an opinion. When I say multiple I mean more than 2 which means its best to way a few more days.

There is no luck needed for me I don't plan on picking up this game at this moment. Plus i'm still waiting on stock and better pricing on the new gpu's so i'm sitting this one out till probably spring or summer. There is no rush for me I can wait.

 

[Conahl]

benderman, Jarred explained to you why the results could be different. unless they list exactly HOW they tested, what settings what hardware etc, you cant compare the results. which it seems they did NOT list exactly what they tested.

Their GPU sequence has almost no NPC's at all unlike the Dubai scene which may explain the big difference if this game uses a lot of CPU cores (which then favors Ryzen of course and has nothing to do with Nvidia).
I don't know if the benchmark sequence you all using is favoring AMD or you got a quick buck for it, but you can believe what you want.

what graphics settings did they use ? low quality ? max quality ? maybe thats why they get better performance then what jarred saw.

I read 2 reviews from reputable sources that confirm my point.

name your " reputable " sources.