Max Payne 3 Performance Tested, Benchmarked

By on June 6, 2012, 12:37 AM

If you've been a gamer for at least a decade, then you will recognize Max Payne as the PC third-person shooter of the early 2000s. Notable for its film noir style and use of the bullet time effect (The Matrix), Max Payne's character went on to surpass anyone's expectations with several console ports, a sequel, and a feature film adaptation in 2008.

Shortly after that, Max Payne 3 was announced and it's been in the works at Rockstar Studios since. The original game was developed by finnish developer Remedy Entertainment and published by the now defunct 3D Realms. Max Payne 3 promises exciting gameplay and impressive graphics, marking the return of bullet time in action sequences, while retaining the shoot-dodge mechanic from previous titles.

It's been hinted that Max Payne 3 will make the most of current high-end PCs, with DirectX 11 tessellation compatibility and "advanced graphics options" available to PC players. So with that in mind let’s take a look at the test setup and then get into the results. Make sure you check out our visual tour comparing DX9 through DX11 graphics.

Read the complete article.




User Comments: 40

Got something to say? Post a comment
TS-56336 TS-56336 said:

Be best to turn off shadows to reduce memory usage. I'm currently using 544MB out of 1024 of my Sapphire 6790 and it's still smooth on DX11. Set it on HDAO, 2X MSAA, 16XX AF.

Guest said:

Interesting to see that there is no big difference from dx9 to 11 .. but there is a big drop in fps. Seem like this is not the title that will push me to make a GPU upgrade. Nothing new, waiting for the next big game .. ;)

Guest said:

I like these articles, but this one is done wrong. You should have used the latest Max Payne drivers,

ForceWare 301.42 WHQL and,

12.6 CAP Beta 1 (for single card & crossfire).

Staff
Steve Steve said:

Be best to turn off shadows to reduce memory usage. I'm currently using 544MB out of 1024 of my Sapphire 6790 and it's still smooth on DX11. Set it on HDAO, 2X MSAA, 16XX AF.

Turn shadows off? Not sure about that, they add so much detail.

I like these articles, but this one is done wrong. You should have used the latest Max Payne drivers,

ForceWare 301.42 WHQL and,

12.6 CAP Beta 1 (for single card & crossfire).

We did use the GeForce 301.42 drivers and the 12.6 CAP drivers do nothing for single card performance in Max Payne 3 so I am not sure how this article is "done wrong".

customcarvin customcarvin said:

lmao.... You guys did a great job here reviewing the game. Personally I don't think anything was "done wrong" but whatever... and I agree with you Steve... No shadows WTF???

I really like it when you guys measure the performance of PC games like this. I'm constantly building budget gaming PCs for people, and articles like these are a great resource. So, THANK YOU!

Have you guys thought about putting together a browser database search for all the statistics you have gathered --kinda like Anand's cpu compare? That would be pretty cool, lots of work, but cool nonetheless.

Guest said:

Thanks for the benchmarks, would it be possible to include in the 16:9 and 16:10 resolutions (ie. 1440x900, 1920x1080?)

Ma_ga said:

it would be nice to see how it's perform on older GPUs like the HD4xxx series

Staff
Julio Franco Julio Franco, TechSpot Editor, said:

We could always test 100 graphics cards, but is it necessary? Not really.

Would it be feasible given the time constraints in publishing this article shortly after the game's launch? Not at all.

But going back to your particular inquiries. We didn't test the Radeon HD 4000 series because those don't support DX11 and we made this a DX11 test. Also, other resolution's performance can be easily inferred, 1920x1080 in particular is almost the same as 1920x1200.

Guest said:

The game runs pretty good for me and my setup is nothing fancy, especially in the GPU dept.

My Specs...

CPU = i3-2120 (I am sure this will beat most of AMD's quad cores even though it's only dual core as in general tests I have seen this CPU beats nearly all (maybe all?) of AMD's CPU's in gaming)

GPU = Radeon 5670 512MB (in NVIDIA terms... it's pretty much almost as fast as Geforce 9800GT or noticeably better than the Geforce 8800 line of cards and it sips power as it runs from the motherboard power)

RAM = 8GB (but anyone with 8GB of RAM will never have issues with games)

Mobo = ASUS P8H61-M LX Plus Rev3.0

I got basically the graphics on 'high' in DX11 mode @ 1920x1080 and my performance is well within playable levels as I don't have trouble aiming although on certain levels with a lot of action on screen with explosions etc the frame rate you can tell takes a noticeable hit but overall it's a pretty solid experience and never comes down to a crawl to where it's any real issue.

I have not checked my exact frame rate but I am sure more often than not it's more than good enough although like I was saying above it does drop somewhat low here and there if the action is pretty heavy but even during those times it never comes to a crawl as like I was saying I can still aim without the frame rate effecting me to any noticeable degree and many times in the game the frame rate is pretty smooth.

Also, once in a while the game will crash/freeze my system (it's did it about 2-3 times so far and once the game itself crashed but it's not bad considering I spent many hours (probably 12+ hours) on it so far) but it's typically not a issue and early in the game I had a pretty big glitch as I could not get threw a doorway which I needed to to get progress. but I just exited the game and reloaded it and it was fine after that. also, the sound is generally in sync but I seen it get out of sync a bit once. but overall the game runs pretty well and no show stopper glitches so far as I am not finished with the game but I am on Chapter 10+ and the game has 14 chapters, so I am through a larger chunk of the game.

Guest said:

Interesting. I'm not sure what I am doing differently, but I am getting FAR higher frame rates than these.

Running my 2500K @4.8Ghz and my GTX 670 Windforce (Gigabyte) at it's stock settings (no power target/gpu offset/mem offset OC'ing), I am getting anywhere from between 150FPS up to 350FPS. Any ideas as to why mine is running so much faster?

Everything is set to the highest it can go with tessellation on very high, HDAO, 16xAF, 4xAA etc, etc @1920x1080.

Guest said:

Ah, it seems that EVGA Precision might not be showing correct FPS which is why mine is showing so high. I will check with a few different programs and see if there is a difference.

TS-56336 TS-56336 said:

Mr. Walton, that's exactly my point. It gave so much detail I had to kill it. Can't imagine an 864 MB usage on my card would make my game lag occasionally. Although, turning shadows off gave me an opportunity to set AF to 16x and 2x MSAA. Question is, would my recommendation suffice? I'm a little uncomfortable with my Radeon 6790 Series because it ain't enough. But I'm making progress.

Guest said:

Ok, I can definitely confirm that EVGA Precision is not so precise at this point in time with Max Payne 3. My Fraps scores are way more realistic at between 80 and 90FPS.

k9182000 said:

The fun part for me is to try to find a copy that is under $89.99 on steam. For some reason Rockstar Studios thinks it's ok to charge Australian's $30 more for the same product. The sad part is the exchange rate at the time of writing this is $1 US dollar = $1.00895 AU dollar. So Rockstar Studios I would like to know why I have to pay $30 more for the same product?? You keep on saying "No one buys our products they just pirate them" this is why they do it. If I look up the US price for a product and it only $5 difference, that I can live with but $30. To make things worst there is NO transport cost to sell the game here in Australia. The funny part is I can import this game for 1/2 of the price on steam. This would be ok if Rockstar Studios supported steamwork (register retail key on steam) but they don't, because they know people will just import. Our government is now looking into why products from over seas cost us so much even when there is zero transport cost.

Here are some examples of the price difference

Max Payne 3 - $59.99USD - $89.99AUD

The Elder Scrolls V: Skyrim - $59.99USD - $89.99AUD

Even a old game

Mafia II Digital Deluxe - $39.99USD - $79.99AUD (wow $40 difference)

All I'm asking for is a fair price is that to much to ask for??

Guest said:

Same guy from posts 11,12 and 14 here. I got my copy from here:

[link]

Get it while it's still cheap. (I'm from Aus as well mate). It's at about $40.60 AUD atm.

Guest said:

I've been playing on a core i5 2500k, gt560 ti sc, 8 gigs of ram at 1920 x 1080 on very high settings, dx11, shadows normal, fxaa high no mxaa and it plays and looks fantastic. I've had no issues at all and it runs super smooth!

totalia said:

my spec

i7 980x

gtx 480

24 gigs ram

resolution 1920x1080, all max settings

getting minimum 90 fps so ya that review done wrong dono how

customcarvin customcarvin said:

my spec

i7 980x

gtx 480

24 gigs ram

resolution 1920x1080, all max settings

getting minimum 90 fps so ya that review done wrong dono how

Umm no offense totalia, but I think you might have done something wrong... What you're saying to us, is that you get better performance than a GTX 680 with your 480... Don't know man, but that does not sound right at all.... Maybe you should recheck your settings and measure with fraps.

LNCPapa LNCPapa said:

<sarcasm>My Q6600 + 8800GT plays it great at max settings and max resolution!</sarcasm>

Worthwhile benchmarking requires consistency and something is obviously different when numbers are that far off. We will never hear the end of people saying they can do better with less... it's just the way things are on the internet.

Guest said:

great article,

but one thing you should really add a Core i3 2xxx to your tests, is a very popular gaming CPU and it adds another interesting CPU to look at, it combines just 2 core BUT 2 cores of higher IPC than the phenoms AND HT,

and while HT might be useless in games like this with 4 core CPUs, it probably isn't with only 2 cores...

fimbles fimbles said:

I get a memory reading of 1792 mb using 2 x 896 mb gtx 275s in sli... strange...

Guest said:

How come the results for 560Ti on exactly the same settings (1920x1200, Very High, DX11) differ by 20% here vs here?

Dawn1113 said:

Well-written and comprehensive review, as always. You guys are the best at this. Thanks for the effort.

I'm not a big fan of third-person shooters. But I must say the graphics on this game look amazing -- and that's just from the screenshots here and the clips I've seen on Youtube so far.

That, plus this review, tells me I should at least give this game a shot. I'm thinking my PC should be able to get 40+FPS from this game, maybe even 50. There are games that are quite playable at those framerates. I'm hoping this is one them.

Thanks.

Guest said:

but one thing you should really add a Core i3 2xxx to your tests, is a very popular gaming CPU and it adds another interesting CPU to look at, it combines just 2 core BUT 2 cores of higher IPC than the phenoms AND HT,

and while HT might be useless in games like this with 4 core CPUs, it probably isn't with only 2 cores...

Indeed, that particular is cheap but pretty fast on games. And about HT. Its not useless. Testing several games with HT enabled and disabled, HT helps a lot for games that are optimized for HT. GTA4 for example, tested Arkham City as well and performance went down when only using 2 threads. Heck even console emulators uses HT and it gives performance benefits

Guest said:

Someone please tell me they still have bullet time.

Guest said:

why no core i3 test ?

Guest said:

Indeed!

core i3 is far more interesting than many of the CPUs tested, it makes no sense!

Guest said:

You can't bench-mark the cut-scenes, they're pre-recorded and played back with Bink.

Staff
Steve Steve said:

Have you played the game? Have you checked the section of the game that we tested? I would do that before you tell us we cannot benchmark it.

Guest said:

I'm running 2xGTX 580 in SLI, and another DX11 card as dedicated PhysX..

I have 4gb GPU mem available, BUT the game eats up ALL of it, even when I'm at lower Quality settings.

is this a Game preoblem, og Nvidia Driver problem, or does the game eat memory like there's no tomorrow?

Guest said:

Why all the complains about not including the Core i3? No one with half a brain would pair a Core i3 with a GTX580/670/680/HD7950/7970 which is what's necessary to max this game out at 1080P with 4xMSAA. Above that resolution, the game is mostly GPU limited. If you have an i3, you are probably using a graphics card that's the limiting component in the first place.

Guest said:

wth are you talking about? they even included Athlon II X2 on this tests, so why not?

the slowest Intel CPU on this test is an i7 920!!

why no Celeron G540 or Pentium Sandy Bridge? and why no i3?

you don't need to play games at max settings,

Zangiry said:

I have 3xGTX 580 in 3-Way SLI, still can't run game at MAX settings :s (I feel that's a bit shocking actually)

And even on lower settings I have Problems with Input Lag, something that can be very frustrating when playing an FPS type game like this.

l2ez4m l2ez4m said:

This is not comprehensive review. Where are ultimate majority of decent budget gaming cpu's - core i3-21xx, any of SandyB Pentiums like popular G860 or G630, Celeron G540, or AMD's apu like A8-3870k. You should definitely update your cpu performance chart to represent probably the bigger half of pc gamers.

Staff
Steve Steve said:

You lost me at this is not a comprehensive review, I challange you to find a more complete Max Payne 3 article anywhere on the Internet. Anyway the CPU results are only a very small part of the testing, we do it simply out of interest sake so see how many cores the game will use. It is primarily about the GPU testing.

Staff
Julio Franco Julio Franco, TechSpot Editor, said:

To complement Steve's comments, not only we run our CPU tests to establish certain patterns (use of cores, scalability of the engine, comparison between AMD and Intel CPUs), but we did test on 16 processors spanning a number of segments. That, along with the other tests make it more data that you will find anywhere on the web for a single game title.

Guest said:

Steve ignore the haters and thank you for a great review please keep it up man.

Guest said:

how would this game perform on my athlon 64x2 4600 with radeon hd5450 ?

Guest said:

Oy.. I like Techspot reviews but some things really get on my nerves. The Athlon X4 645 provided enough oomph the 'exploit' the 680 yet it was called "gutless". Also, the 680 does not outperform the 7970 with 'ease'. It may post better frames in this particular game but although no glitches were reported here, it is well known that there are AMD driver conflicts with this game that Rockstar blames on AMD. Basically, keep up the good work guys but why don't you cut the anti-AMD BS. And Rockstar, quit ripping people off over seas.

Staff
Steve Steve said:

It was called gutless because it is, the Athlon II processors are just about the most gutless desktop models on the market right now. The GTX 680 was 24% faster at 2560x1600, forgive us for calling that an easy win. Those are the kind of performance gains you hope to see from a next generation card so I hate to know what your idea of an easy win is. Finally we love AMD, if stating the truth makes us anti-AMD then I am sorry.

Load all comments...

Add New Comment

TechSpot Members
Login or sign up for free,
it takes about 30 seconds.
You may also...
Get complete access to the TechSpot community. Join thousands of technology enthusiasts that contribute and share knowledge in our forum. Get a private inbox, upload your own photo gallery and more.