Hitman: Absolution - CPU and GPU Benchmarks (GameGPU.ru)

[aaksheytalwar]

Absolution is DA GAME

I can play at these settings:

1080p maxed out 4x MSAA
Minimum FPS > 60 even with VSync locked at 60 I guess. Avg is much higher.

1440p maxed out no AA
Minimum FPS > 60 and VSync locked at 60, hasn't dipped till now.

Very well optimized.

 

[thilanliyan]

Good to hear that the game is good. Will wait for a steam Christmas sale though since my backlog is HUGE! Haven't even played metro 2033 or arkham city yet!! Going through War in the North right now.

 

[aaksheytalwar]

I don't think metro or any batman game comes close to this.

 

[Souv]

metro 2033 and hitman absolution are the two most graphically demanding games out there...
see another bench,amd is rocking with their legendary never settle driver 12.11 (beta8):

http://www.techspot.com/review/608-/page2.html:whiste:

 

[aaksheytalwar]

Hitman Absolution isn't demanding at all. Metro is 2-3x as demanding.

 

[dust]

I don't think metro or any batman game comes close to this.

Lol!

Don't know about batman games, as I never played any, but the stealth is horrible in Absolution compared to Metro(if you ever played the front-line level in Metro without killing anyone you will know this to be true). There are so many dumb things in this Hitman title, it's not even funny.

Not exactly spoilers but:

Spoiler

Aside from the fact that a body being spotted only causes a 30 seconds adrenaline rush for the nearby guards, why do they even bother to report in any weird noise happening around them if no one even cares to check on them again. You can enter any garbage bin if anyone spotted you, but you can safely wait inside(and peek) until the guard(s) pointing the guns at you from right next to your bin, get bored and go back to their routine.

You can literally "hide" at a laptop in the middle of a full room(with the right disguise) and everyone is fine with it. Loose the "hiding" place and everyone sees you all of a sudden.

Subdue an enemy and get reduced point deduction, "hide" the body over the ledge, from 20 meters height, killing him in the process of course, but it's all fair game, no further penalty applies.

I wish the next one will have less scripted guard paths also.

 

[aaksheytalwar]

I don't like stealth games any way

 

[Souv]

without even fully maxing out ,but only at 2xaa and 8xaf @ 1920*1200 only 7970 ghz edition managed to get 61 fps,7970=55,7950=51,gtx 680=49 fps at hitman absolution
........u call that not demanding...haha
........whereas, at full maxed out dx11 metro gets 53 fps on 7970
.....here is the link-http://www.legitreviews.com/article/2059/7/

....metro and hitman absolution are equally and most demanding games dude...just looking at the benches may be absolution surpassed metro on graphical demand....

therefore devs said-

http://www.videogamer.com/xbox360/h...ution_assets_were_built_for_high-end_pcs.html

http://www.pcgamer.com/2012/08/14/h...t-looks-ten-times-more-awesome-on-pc/:whiste:

 

[ICDP]

Lol!

Don't know about batman games, as I never played any, but the stealth is horrible in Absolution compared to Metro(if you ever played the front-line level in Metro without killing anyone you will know this to be true). There are so many dumb things in this Hitman title, it's not even funny.

Not exactly spoilers but:

Spoiler

Aside from the fact that a body being spotted only causes a 30 seconds adrenaline rush for the nearby guards, why do they even bother to report in any weird noise happening around them if no one even cares to check on them again. You can enter any garbage bin if anyone spotted you, but you can safely wait inside(and peek) until the guard(s) pointing the guns at you from right next to your bin, get bored and go back to their routine.

You can literally "hide" at a laptop in the middle of a full room(with the right disguise) and everyone is fine with it. Loose the "hiding" place and everyone sees you all of a sudden.

Subdue an enemy and get reduced point deduction, "hide" the body over the ledge, from 20 meters height, killing him in the process of course, but it's all fair game, no further penalty applies.

I wish the next one will have less scripted guard paths also.

I noticed some of these issues as well and some are very annoying. Unfortunately developers have to make concessions for game play purposes. If enemies in games were as realistic as real humans you would die constantly. If the AI had to do a redio check in every 30 seconds you would never complete a level, because in real life if that happens the target is moved.

The hide body over the edge thing comes up as non-target enemies killed for me. I had to restart a level because I thought I was being smart but the game treated these disposals of unconscious bodies as deaths.

 

[aaksheytalwar]

without even fully maxing out ,but only at 2xaa and 8xaf @ 1920*1200 only 7970 ghz edition managed to get 61 fps,7970=55,7950=51,gtx 680=49 fps at hitman absolution
........u call that not demanding...haha
........whereas, at full maxed out dx11 metro gets 53 fps on 7970
.....here is the link-http://www.legitreviews.com/article/2059/7/

....metro and hitman absolution are equally and most demanding games dude...just looking at the benches may be absolution surpassed metro on graphical demand....

therefore devs said-

http://www.videogamer.com/xbox360/h...ution_assets_were_built_for_high-end_pcs.html

http://www.pcgamer.com/2012/08/14/h...t-looks-ten-times-more-awesome-on-pc/:whiste:

I am sorry but you will need to find somebody with more time to deal with a complete noob like yourself

You don't own either of the games, nor any of the cards, while I do. So I really don't need to waste time on theory. Long story short, buy a 7970 ghz and those games and then tell me your experience.

 

[aaksheytalwar]

Fine. I will try to explain once.

1. In game benchmarks mean NOTHING. Some games play get much higher fps than built in benchmarks, while some perform much worse. What you get in game real time is what matters. Not a recorded demo benchmark which is rarely accurate.

2. I own a 7970 at ghz plus oced speed. I can play it with 60+ fps minimum at 1440p no AA or 1080p 4x msaa. Either way everything maxed out. And I still get 70-100+ fps avg.

3 with the same setup, I don't high that high fps at 1080p high (not ultra) no AA with metro.

4. In short, review websites are crap. Buy the card. Buy the game. And test yourself.

5. Reviews are only relative guidelines, not accurate experiments.

 

[DiogoDX]

I have to desagree. The bechmarks serve their purpose very well which is to compare the performance between video cards. Is logical that 1-3 minute tests will not represents the entire performance of hours of campaign.

And I like when tests are used with much effects Metro and Hitman benchs because you can see how the cards behave in heavier situations.

 

[AtenRa]

The build in benchmark is meant to test your system and give you an idea of what you will get in the game. There are areas in the game that will get you higher fps than the build in benchmark and areas where you will get lower fps.

 

[boxleitnerb]

Built in benchmarks are of little value usually. Shogun 2, GTA 4, Metro 2033 are good examples for massive deviation between these benchmarks and actual performance in the game.

The best way to benchmark games is to actually play them and record with fraps. It's more work to do it reliably, but it's almost always more representative.

 

[ICDP]

The Hitman in game benchmark uses the Chinatown level with a massive number of fully rendered people. It is in no way indicative of the actual game play performance. Most of the game is spent in smaller, indoor levels with a handful of people in view at most.

Performance in the game is better than this canned benchmark designed to bring GPUs to their needs.

 

[DiogoDX]

Built in benchmarks are of little value usually. Shogun 2, GTA 4, Metro 2033 are good examples for massive deviation between these benchmarks and actual performance in the game.

The best way to benchmark games is to actually play them and record with fraps. It's more work to do it reliably, but it's almost always more representative.

Unless you play the whole game the result will be the same as the internal bench. It will only indicate the performance in that area of the test.

I generally prefer the internal benchs because they usually has effects that are present in the entire game and a arbitrary run with fraps might end up losing some of these effects. They also have the advantage of easy reproduction.

 

[boxleitnerb]

Unless you play the whole game the result will be the same as the internal bench. It will only indicate the performance in that area of the test.

I generally prefer the internal benchs because they usually has effects that are present in the entire game and a arbitrary run with fraps might end up losing some of these effects. They also have the advantage of easy reproduction.

The reviewer has to know the game somewhat of course and select a proper scene for the benchmark. It should by no means be arbitrary but demanding and representative.

That way you can be sure of what you benchmark. Lazy reviewers may not know the game at all but just run the benchmark and that's it. For example in Shogun 2 and GTA 4, the benchmarks provide much higher fps than what you actually experience in the game. Or Far Cry 2 where mostly the "Ranch" benchmark is used while that doesn't represent the more demanding fighting scenes. Or people benching Diablo 3 early on while later the game becomes much more demanding with area of effect spells, monster groups and possibly multiple human companions. People have to know the game to benchmark it, there is no way around that.

I can't stand these easy solutions, because they often are sloppy and full of errors. Benchmarking is not easy, it's an art that is not easily mastered.

 

[Souv]

those are all in-game playing benchmarks in absolution....
.....not timed games in-built benches ....

if a single 7970 GHz(single best soc card) plays the game well then the game is not demanding??!!!!!....

that game is strangling every other gpu out there without even full maxed out on AA and AF except uber high end models as 7900 series or gtx 680/670....that's how demanding the game really is....

guru 3d or tech spot has done the bench with in-game playing bench not cayman test:

http://www.guru3d.com/articles_page...phics_vga_performance_review_benchmark,5.html

the game gets low fps except high end models.....

(and if you have a 7970 ghz then it's great news...i appreciate it...but don't defend it saying the game is not demanding as i can play on the fastest single gpu in market!!and u also made a hell out of OC of ur 3770k and hd 7970......and also don't be emotional saying me a complete noob in an argument,then you are loosing your judgement)

 

[aaksheytalwar]

You are a complete noob. I rest my case. It is a waste of time talking to noobs.

 

[2is]

Staying up to play this tonight. Already have it preloaded. It's been six years since Blood Money, way too long in my book.

Tessellation makes sense for 47's iconic dome, considering you stare at the back of it the whole game. Looks like I'm going to have to crank the clocks on my 7970 to get decent frame rates @ 2560x1600.

Looking at those benchmarks, I hope you mean 7970's If you're cranking up the settings anyway.

"but can it run hitman" just doesn't sound right though...

 

[aaksheytalwar]

A single 7970 ghz can max out this game at 1440p without any AA.

 

[OVerLoRDI]

Performance on my system is pretty rough sadly. I think there might be something weird going on with my drivers (might be time for a reformat)

Getting about 55 fps average on Ultra settings with MSAA 4x @ 2560x1440

7970 crossfire @ 1225mhz
i5-3570k @ 4.5ghz

 

[paladinm]

Bite me. I think the new Hitman doesn't looks THAT good. I mean, it doesn't gives the the shock when I first saw Max Payne, DOOM3/Half-life2, Farcry1, or Crysis1. Plus, I see no significant improvements in graphic, comparing with the other latest games.
So I don't know why it's such graphic-card-consuming

 

[Jaydip]

The game does look good but the stupid checkpoint system is horrid.

 

[AdamK47]

Bumping this thread. Just got the game today. It looks spectacular and seems like it would be demanding. I didn't know it had an in-game benchmark. Decided to run it and post results.

Specs:
Intel Core i7 3960X @ 4.5GHz
ASRock Extreme11 @ 36 x 125MHz
32GB G.Skill Ripjaws Z @ 2333 DDR
Two Nvidia GTX 690s in Quad SLI
-- 310.70 drivers / driver defaults / +85 core +500 memory
256GB Vertex 4 SSD
Six 512GB Vertex 4 SSDs in RAID-0
4TB Deskstar 7K4000 HDD
Pioneer BDR-206 BD-RW
Cooler Master HAF-X case
Thermaltake Water 2.0 Extreme cooler
Corsair AX1200 power supply

Settings:

All results are at 2560x1600.

No AA:

4X AA:

8X AA:

The minimum framerate seems to be from the transition from when the benchmark stops to when the results are posted. It doesn't show it at all during benchmarking.