Call of Duty: Black Ops GPU & CPU Performance In-depth

By on November 16, 2010, 4:20 AM
Two years ago we published an in-depth performance review of the fifth major installment in the Call of Duty series, World at War. Since then Call of Duty: Modern Warfare 2 was released almost exactly a year ago. Now on to a successive and very successful yearly release, Call of Duty: Black Ops arrives, paving the way for even more games in the series.

Developed by Treyarch, Call of Duty: Black Ops debuts a new theme around the Cold War conflict. We test 19 graphics card configurations, using AMD and Nvidia GPUs that ranged from the uber expensive models to budget-minded offerings.


Besides the fact that it's coming from the same development team, Black Ops essentially uses the older Call of Duty 4 game engine. To be precise the game runs on an enhanced World at War engine, which is in itself an improved version of what Call of Duty 4 used back in 2008. This new revision of the engine features a streaming texture technology also supported in Modern Warfare 2 that allows for larger levels, for example the "Payback" level where the player controls a helicopter. Additionally, lighting effects have been improved and the game supports 3D imaging rendered when using the correct hardware.

Read the complete performance review.




User Comments: 76

Got something to say? Post a comment
blimp01 said:

Great review this is what I've been waiting for lol.

I wouldn't think that call of duty would ever be this CPU intensive, my old 775 struggles in this game.

pgcharlie said:

Sir, u mentioned under cpu performance that i3 540 crashed and burned.. u literally mean burned or was it a technical term?.. i own that processor with Asus H55 mobo with 2 GB RAM, 1 GB Radeon 5700 series card...n i agree with Blimp01.. great review.. thanks

Guest said:

Any chance in getting a SLI / Xfire scaling review for this as well? would be nice to see the numbers.

Staff
Julio Franco Julio Franco, TechSpot Editor, said:

I'd dare to say multi-GPU testing is not really needed with the data you have at your disposal right now. Any graphics card at the level of a GeForce GTX 460 or better will perform solidly when coupled with a quad core or even triple core CPU.

We can also confirm Crossfire and SLI work just fine on Black Ops, so you will see proper scaling as in other games that are well optimized. Check out these two articles for such scenarios:

[link]

[link]

SilverCider said:

Ah, makes me happy that I went down the i5 750 route early this year XD. Great review btw :P

whiteandnerdy said:

I have an i3 540 and i had no trouble getting 50-60 FPS through the entire game (i play with fraps on all the time). My 5850 may have helped a little but i didn't think it was bad. Great game. Great review. thanks

bugejakurt said:

I have a Core 2 Duo E8300 (2.83 Ghz) CPU and it totally freaks out with COD: Black Ops. My 9800GT renders it easily but my CPU is constantly 80-100%. Sometimes it reaches 100% and begin lagging badly. I don't know why but in my opinion I don't have a bad CPU I don't know why theses spikes occur in some situations. Anyone else with this problem? Anyone suggesting a fix?

LNCPapa LNCPapa said:

This review made me very happy - just so happened that you used hardware identical to mine and it produced some of the best numbers.

Guest said:

There's something very wrong with you HD 4890, those numbers can't be right!

Ranger12 Ranger12 said:

A 9800gt and a 4850 but no 4870? Guess I'll have to extrapolate. Or maybe that's a sign I need a new card....

Guest said:

This performance review is completely premature. There are several known GPU and CPU hitching issues, which would result in inaccurate numbers across the board. I'm disappointed I didn't see more of an emphasis put on this. This performance review will misinform the many people who come here to research what hardware they should buy under the false pretense of a broken game. I suggest either noting that more prevalently in the review or waiting till all of the bugs/gltiches are ironed out for more hardware profiles before you provide such information.

grvalderrama said:

Hey, Guest, bugs have nothing to do with the general performance/requirements of the game. If there were bugs in this review, which I don't believe, you will see a slightly increase in your performance once those bugs are fixed.

pretzelwagon said:

I would agree with the latest comment. While the game's story is pretty good, and it stays true to the classic Call of Duty style, there are some huge FPS (frames per second, not first person shooter) stutter issues. These issues have been addressed for some platforms (allegedly), but I still have freeze-up issues where the screen just drops from a V-synced 60 FPS to 0 for a few seconds. A little disappointed this article didn't touch on that more, and then update later, much like was done with Splinter Cell: Conviction here at TechSpot. Good info, though.

princeton princeton said:

grvalderrama said:

Hey, Guest, bugs have nothing to do with the general performance/requirements of the game. If there were bugs in this review, which I don't believe, you will see a slightly increase in your performance once those bugs are fixed.

Not exactly. This is the type of fucked up game where 2 of the same exact system may have noticeable differentials in performance. At the moment the game on PC doesn't have any bugs. It's a bloody bug to begin with.

pretzelwagon said:

Sorry grvalderrama, your comment hadn't shown up when I posted. I was referring to guest's comment. And typically I would agree with you, but these bugs are so detrimentally attached to the frames per second, and only seems to hit certain platforms and CPU/GPU combinations, that I wonder if it actually does change the average FPS for certain cards/CPUs. Regardless of the overarching effects, completely overlooking this issue is surprising to me.

Guest said:

This game run great on this setting with my Athlon x2 3.0GHz with unlocked Radeon HD4830 :)

TomSEA TomSEA, TechSpot Chancellor, said:

Asking for a game not to be performance reviewed until all the bugs are fixed is ridiculous. People are buying and playing the game NOW. A performance review is perfectly in order. You could wait for years for all the bugs to be worked out of a game to do a review - what good would it be then?

Guest said:

This isn't a black and white situation. It's utterly ridiculous that no mention of the game breaking GPU hitching problem wasn't mentioned at all in a performance review. Again, hardware performance based on false pretenses of a broken game. It's an injustice to anyone who doesn't know the wiser and reads this misinforming review and an injustice to the hardware manufacturers who are being misrepresented by a bug ridden game. And we're not talking minor bugs, we're talking about bugs that DIRECTLY AFFECT FPS in a negative way.

Here's an example of spreading mass misinformation, from the review: "Although Black Ops has been out for just a few days, graphics card manufacturers already got their multi-GPU technologies working perfectly with it. This is one advantage of using a tried and true game engine. "

PERFECTLY?! For shame.

Signed,

Elessar (I made the first complaint a few posts back)

Guest said:

FIrstly TOP REVIEW.

However, I have to agree with the above user. There is a 43 page thread on the OCUK forums regarding the massive inconsistencies in performance. For instance, I have a QX9650 @ 4Ghz + 5970 @ 900/1200 and I am getting massive freezes in game. We have actually had reports of users with overclocked i7's running on Quad Fire 5970's getting the same performance. This is not micro-stuttering people, it's serious LAG. If you don't believe me refer to @pcdev on twitter.

So on that note, I'm not going to rush out and buy a i7 950 just for this game - yet. I belive there is a MAJOR performance patch coming out either this weekend or the next so I think it's best to wait.

To the author, is there any chance you could re-run your tests once that update is out?

Guest said:

"So on that note, I'm not going to rush out and buy a i7 950 just for this game - yet."

This is exactly why this performance review is more than just misinforming. It's flat out detrimental. The uninformed may read this and go out to buy hardware they THINK may run this game better, when in reality it's not their hardware at all. The above poster is smart to wait for a performance patch, but no mention of any of this in the main article is just terrible and may very well end up wasting people's money.

LNCPapa LNCPapa said:

Perhaps these graphical hangs or fps drops to 0 were not experienced during the benchmarking process of this review - could that be the case, Steve?

Guest said:

I've been having the same issues Elessar (Guest) has, and I've read many posts on other forums that this is a known issue, for the PC version in particular. The stuttering and FPS slowdown are game-breaking for me, I can't play it at all. My rig is only 19 months old, I'm running a Core2 Duo E8400/3GHz CPU, NVIDIA GeForce 9800 GT (512MB), with 4GB RAM. There is no reason my rig shouldn't be able to run a game built on a souped-up Quake 3-engine, and run it well. This is a known issue that is not impacted by lowering graphical settings, though I obviously tried that at first.

I'm not saying that you shouldn't do a performance review of a game until all the bugs are fixed--as someone rightly pointed out, that would mean no performance reviews would ever be written. But a performance review is the one place I would absolutely expect to see mention of a widespread performance-crushing bug, and I see no mention of it here.

Guest said:

Excuse me as I am not "pc requirement savy", however I tested mine against Activisions requirement test and gamesystemrequirement.com and both said I could run the game. Unfortunately like so many others I am unable to play, much less get past the "Press the key" screen as I have crashed with the "blue halo" of death over the cursor.

Hopefully awaiting a patch/fix. Ironically I can play Medal of Honor Tier 1, go figure.

Vista 32

AMD PHENOM 9750 QUAD

ATI 3650

Guest said:

Ha running Amd x64 athlon 5200+ and gts 250 gpu i get 47-120 fps so how are these stats right they cant be and the game blows on performance and networking why not write a review about that oh forgot not paid to tell the truth.

Guest said:

Would be nice if you could toss an q9650 in the mix!

KG363 KG363 said:

Thank you for including an X3 processor after asking many times.

Keep it up!

And what was wrong with the 4890? It's fps was miserable

Guest said:

I actually have an E8500 running overclocked @ 3.6 per core, with an HD 4890, and 8 GB of DDR2 on Vista 64-bit. I'm getting horrible performance on this game despite being well over the requirements. It really needs to be fixed badly on the part of the developer. It would appear the game was made for consoles, and PC was some sort of afterthought unfortunately :/

Johny47 said:

Very nice, thanks for doing this as always. And at 2560x1600 you really don't need 4x anti aliasing you show off's =P

unless you like stopping and staring at leaves and debris hahaha =)

Guest said:

I'm not sure how this accounts for the issues that I, and many others are having.

specs: e8400, 4gigs ram, GTX460 1 GB on a 1080p monitor

My computer meets recommended settings though still maintains 90-100% CPU usage at all times regardless of graphics settings. I'm currently playing on lowest settings, no AA, shadows etc... and continue to experience crippling lag.

BlindObject said:

GTX465 (which you missed...again)

Q9550 at 3.5ghz

4GB DDR2 at 945mhz

Eats the game with 1080p maxed out >~60fps easily.

Why do you keep missing the GTX465? It might not be popular..but some us do own it. We just haven't flashed to 470 yet...

Staff
Steve Steve said:

LNCPapa said:

Perhaps these graphical hangs or fps drops to 0 were not experienced during the benchmarking process of this review - could that be the case, Steve?

Nice to see someone still uses their head When testing we ran into no problems at all. I have finished the single player campaign and did not run into a single bug. I have no doubt there are bugs with the game, though they never affected us. All cards and configurations benchmarked without a hitch with the exception of dual-core processors which run like crap due to being over taxed.

Either this game is too demanding for dual-core processors or there is a bug, in either case you cannot run this game using the highest visual quality settings on a dual-core, at least from what I saw.

kg363 said:

Thank you for including an X3 processor after asking many times.

Keep it up!

And what was wrong with the 4890? It's fps was miserable

Hey we only introduced the triple-cores this time because it made sense, the dual-cores didn't perform well and the quad's did so we wanted to check the middle ground

Also there was a typo with the 4850 and 4890 cards, they were mixed up, I have double checked the excel spread sheet and I made a copy and paste error sorry. I have double checked all the results now and everything is in order. Thanks.

BlindObject said:

GTX465 (which you missed...again)

Q9550 at 3.5ghz

4GB DDR2 at 945mhz

Eats the game with 1080p maxed out >~60fps easily.

Why do you keep missing the GTX465? It might not be popular..but some us do own it. We just haven't flashed to 470 yet...

You have a quad-core so I am not surprised by your performance. As for the GeForce GTX 465 I have addressed this time and time again. First of all if we had a card we would not include it as it's not a critical release. Given that we include the GTX 470 and GTX 460 you can work out where the GTX 465 would be. The GTX 465 is a card that should have never been released. Finally we were never sent one because ... well they suck and no one wanted to put their name to it in the end.

I've been having the same issues Elessar (Guest) has, and I've read many posts on other forums that this is a known issue, for the PC version in particular. The stuttering and FPS slowdown are game-breaking for me, I can't play it at all. My rig is only 19 months old, I'm running a Core2 Duo E8400/3GHz CPU, NVIDIA GeForce 9800 GT (512MB), with 4GB RAM. There is no reason my rig shouldn't be able to run a game built on a souped-up Quake 3-engine, and run it well. This is a known issue that is not impacted by lowering graphical settings, though I obviously tried that at first.

I'm not saying that you shouldn't do a performance review of a game until all the bugs are fixed--as someone rightly pointed out, that would mean no performance reviews would ever be written. But a performance review is the one place I would absolutely expect to see mention of a widespread performance-crushing bug, and I see no mention of it here.

Why are you "guests" convinced this is a bug? We saw high utilization across 4 cores, needless to say a dual-core processor is not going to like this. You have an old dual-core processor and are experiencing the lag we saw, again why is this a bug? Get a quad-core and see if the bug goes away?

This is not the first game to run terribly on a dual-core, Battlefield Bad Company 2 does as well.

Guest said:

I have a CORE2 quad 2.66 o/c by 20% and a GTX460 running of Windows 7 and it runs like a sack of crap with this game. It's only just been rebuilt and flattened and should be able to run no problems but I often see the CPU running flat out, game freezes and anywhere from 20-90FPS when I say 90 it's not often and usually for a couple of seconds. There are major bugs with this game and in my opinion it's a no brainer don't buy the game your wasting your money till they fix the problems and don't by a server like I did, as I'm now paying to play a buggy/stuttering game with a poorly design Rcon and very little you can program! If you just want to sit there kicking people all day then the server is the way however if you want to buy a clan server and play as a clan or your mates don't bother.

Guest said:

I believe many people responding have failed to realize that these tests were only in campaign mode. Maybe there needs to be a test of multiplayer mode as well?

Guest said:

please i am begging you next time add a Q9650 too :(

Staff
Steve Steve said:

Guest said:

I believe many people responding have failed to realize that these tests were only in campaign mode. Maybe there needs to be a test of multiplayer mode as well?

It is difficult to accurately test the multiplayer mode without times demos. I agree though I think most do not realize its single player only performance.

klepto12 klepto12, TechSpot Paladin, said:

i love how all the guests hide behind there no names and ***** at a guy who worked his ass off to give us this review give him some respect all the time he spent to do this is probably more than some of you work in a week.

LinkedKube LinkedKube, TechSpot Project Baby, said:

Man, you have to have thick skin to work at TS. Nice review. Just got back from the weekend. Now off to crysis to run test for someone.

Definately gonna pick up black ops tomorrow. Looks good.

klepto12 klepto12, TechSpot Paladin, said:

Man, you have to have thick skin to work at TS. Nice review. Just got back from the weekend. Now off to crysis to run test for someone.

Definately gonna pick up black ops tomorrow. Looks good.

i heard black ops is nice haven't got it yet but supposedly even the most amazing systems have stuttering problems while playing must be a bug or something bu smash your system could chew this game up and spit it out like nothing

UT66 said:

this game is garbage, i couldn't even finish it, in tired of the same old boring shooting gameplay. the good news? the patch fixes the performance issues ( works great on a my E6600 + 4770) but at 1080p this is butt ugly to me, the shadows just looks better at 720p because of their lower resolution. it makes no difference anyway, like y said, garbage.

UT66 said:

"Get a quad-core" i see so that's the point of this article? sorry but no, i rather buy better game- there is nothing wrong with my core duo, in fact my duo crushes both the "cell" and the garbage of powerpc inside the 360. this is a console game after all, even a calculator should be able to run it at 60fps.

fpsgamerJR62 said:

One of XP SP2 PCs has a 2.93 Ghz Core 2 Duo E7500 CPU, 2 Gigs of DDR2 memory, a 3-year old 512MB GF 8800GT card and a Viewsonic VA912 LCD non-wide19-inch monitor which tops out at a resolution of 1280x1024. I wonder how well Black Ops will run on this rig. Every other PC seems to running at 1680x1050.

Guest said:

i had the same blue screen crash at first just updated my nvidia 9800gt drivers and the game ran fine

Guest said:

All of your test are done with directX 9 i am curiuos why? Every machine capable of running the game would most likely have directX 10 in most cases and possibly directX 11 as is my case.

Guest said:

Very interesting review

I'm running a i7 860 @ 3.4Ghz with an 5970 at 5870 speeds.

I'm having major issues with crossfire scaling :( With catalyst AI (crossfire) enabled for the 5970 i will have a framerate of 50-70fps (com_maxfps = 125) with catalyst AI disabled i will hit 80-125 fps.

Ingame settings are all set to lowest and some fancy stuff in the config has been turned off, the resolution used is 1680x1050.

This shows to me that somehow crossfire performance scaling is negative. So what i would really like to know is why the issue does not arise in your testing. Which Catalyst drivers are you using? I have tried 10.5 through 10.11 with no notable difference.

Staff
Steve Steve said:

All of your test are done with directX 9 i am curiuos why? Every machine capable of running the game would most likely have directX 10 in most cases and possibly directX 11 as is my case.

You should test it in DX11 and let us know how you go

red1776 red1776, Omnipotent Ruler of the Universe, said:

, post: 966145"]You should test it in DX11 and let us know how you go

Guest said:

This review is not reflective of what I'm experiencing in game. And from what I'm reading on various Black Ops forums, there is a large percent of players in my boat. One week after release, game is still not playable.

I'm averaging around 20fps in MP at 1680x1050 (no noticable change at 1280x1024)

my comp:

Phenom II X4 940 @ 3.00 Ghz

4.00 GB RAM

480 GTX w/ 1.5GB RAM

Guest said:

This review was a surprise to see with how buggy this game is. It will need an update after the game has been patched for sure.

Staff
Steve Steve said:

This reveiew is not reflective of what I'm experiencing in game. And from what I'm reading on various Black Ops forums, there is a large percent of players in my boat. One week after release, game is still not playable.

I'm averaging around 20fps in MP at 1680x1050 (no noticable change at 1280x1024)

my comp:

Phenom II X4 940 @ 3.00 Ghz

4.00 GB RAM

480 GTX w/ 1.5GB RAM

This review also does not reflect MP performance in anyway...

This review was a surprise to see with how buggy this game is. It will need an update after the game has been patched for sure.

If they update and change the single player performance we will update for sure.

Guest said:

This is the second well-threaded game in a row where the numbers for the i3-540 seem low. The last game MOH also showed the same thing with the i3-540 no faster than the G6950 clock for clock. You have me convinced something is wrong with your i3-540 setup. Someone needs to bring up the Windows Task Manager (CTR+ALT+DEL) and click the performance Tab to check to make sure all 4 logical processors are functioning. My bet is hyperthreading is turned off and your I3-540 will only show 2 of the 4 logical cores active.

In my last comments I stated that the i3/i5 dual cores can see a 50% boost in frame rates with HT, as HT does generally very well in well threaded games. The quad cores with HT (i7) don't seem show as large an increase because the 4 physical cores appear to take care of most of the load.

A few examples of this i3 HT performance.

RE5

i3-530 ... 153.7 ... ... 65% faster

G6950 ... 93.0

Colin McRae : Dirt 2

I3-530 ... 84.69 ... 50.3% faster

G6950 ... 56.35

Source ... http://www.computerbase.de/artikel/prozessoren/2010/test-int
l-pentium-g6950-und-core-i3-530/23/#abschnitt_resident_evi
_5

DAO (Dragon Age Origins)

i3-530 ... 95.0... 56% faster

G6950 ... 60.9

Source ... http://www.anandtech.com/bench/CPU/60

BF - BC2 multiplayer

See ... http://forums.overclockers.com.au/showthread.php?p=11398817#
ost11398817

About half way down the page of the link is a graph of the i3-530 with HT off and on. The frame rate increase (an eyeball estimate) goes from an average of just under 40 to close to 60 -- about a 50% increase. Even more significant is that the minimum frame rate goes from dips into the low 20's (several times) to just one dip below 40 (mid-high 30's) -- a even greater boost than 50%.

Your i3-540 just doesn't seem to be performing where I would expect to see it in multi-threaded games.

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.