Tomb Raider Tested, Benchmarked

Really enjoying the game so far and actually quite surprised at the excellent day 1 AMD support - Crossfire scaling is good (60-80%) and there's no microstutter. TressFX is just way too much of a performance hit though, hard to justify for just the hair.

Btw if anyone is having problems with CF not working, make sure Exclusive Fullscreen is turned on, because turning it off makes it run in windowed mode I think (and CF doesn't work in windowed mode).
 
I think the 6GB of GDDR5 is the saving grace for the Titan haha. The drivers arent mature yet, but that extra memory allows it to perform to the AMD level despite badly optimized drivers.
@slh, I can image that it will be a performance hit, but some people seem to like nice hair haha.

As for the CPU benchmark, surprising to see that the game doesnt improve with a higher clock speed. It seems to be very GPU intensive.
 
Has anyone tried this game with a GTX690 please? (I'm wavering between upgrading to the 690 or the new Titan)
 
The test is wrong
with the new driver 314,14 and the new patch ... I 680 ultimate sttings 1200p 55fps

fail.........test

Although that is a silly comment to make you did spot a typo, we did test with the 314.14 drivers. As for the comment about yesterdays patch get real please.
Would it be possible to update a few of the Nvidia benchmarks since the game was patched only yesterday? I had a friend that bought this game and couldn't even run it with his 560ti and I assured him it would work after the patch, but I have absolutely no idea what the patch has done for framerates. I didn't buy the game but I'd be able to run it with 680s in sli anyway.

Has anyone tried this game with a GTX690 please? (I'm wavering between upgrading to the 690 or the new Titan)
Multiply the 680 by about 1.8 and you'll get the 690's framerate pending the scaling is complete garbage.
 
Sorry the test is wrong

Kappaalphataualphagammarhoalphaphi3AE0-60_zps1049a351.png


Kappaalphataualphagammarhoalphaphi3AE0-60_zpsdc5be4e1.jpg


Kappaalphataualphagammarhoalphaphi3AE0-60_zpsc7e0445a.jpg





Kappaalphataualphagammarhoalphaphi3AE0-60_zps5990afbe.jpg
 
Sorry the test is wrong

Sorry no it isn't. It looks like those results are with the new patch which just came out so obviously we have not had a chance to re-test just yet.

Also don't forget we are testing a different part of the game which is very different to the built in benchmark.

"For now, those wanting to play Tomb Raider are far better off with an AMD solution as the HD 7970 GHz Edition was able to deliver more consistent performance than the GTX Titan and it offered substantially better results than the GTX 680, which ranked lower than the HD 7870."

Not so fast...
http://www.fudzilla.com/home/item/30686-pc-tomb-raider’s-performance-issues-being-looked-into

Not so fast? You understand what "For now..." means right because that is how we started that paragraph.
 
Intel Q9300 @ 3 GHz
Windows 7 x64
Sapphire Radeon 7870 XT
Tomb Raider build 1.0.718.4 built-in benchmark
1680x1050
Quality,Min,Max,Avg
Normal,102.0,160.0,139.2
High,80.0,114.0,98.2
Ultra,54.0,78.0,67.3
Ultimate,33.9,56.0,45.3
 
In saying the above though, I'd like to clarify.. On the screenshots showing graphical presets, Ultra has tressFX to normal, while Ultimate has it set to On. If we look at the % difference in performance on the two presets, the NVIDIA cards get absolutely demolished on Ultimate.

I was under the impression that TressFX is AMD only at this stage? Also, if we look at CPU performance, there is no difference when OCing using Ultimate, therefore the physics isn't being processed on the CPU.

I'd gather that the TressFX doesn't work on NVIDIA properly, and this is having a large impact on performance. I don't know what other presets for hair there is, apart from normal/on, so I can't confirm this as much as I'd like.
 
In saying the above though, I'd like to clarify.. On the screenshots showing graphical presets, Ultra has tressFX to normal, while Ultimate has it set to On. If we look at the % difference in performance on the two presets, the NVIDIA cards get absolutely demolished on Ultimate.

I was under the impression that TressFX is AMD only at this stage? Also, if we look at CPU performance, there is no difference when OCing using Ultimate, therefore the physics isn't being processed on the CPU.

I'd gather that the TressFX doesn't work on NVIDIA properly, and this is having a large impact on performance. I don't know what other presets for hair there is, apart from normal/on, so I can't confirm this as much as I'd like.
knowing nvidia, they will go after tressfx and bring those features to CUDA. They already have hair that can be computed with CUDA, but it isnt implemented in games, but more in animated films
 
Is it possible for me to get a considerable framerate while TressFX is enabled? I have the Radeon 7770, but in Normal settings.
 
Is it possible for me to get a considerable framerate while TressFX is enabled? I have the Radeon 7770, but in Normal settings.


I doubt it. TressFX is very graphics intensive. I can play maxed out with TressFX off and have a enjoyable experience. But if I turn everything to normal and 8xAF w/ TressFX on it gets extremely choppy. Granted I'm only using a 6970 and on a massive resolution but the impact is just so great. It's more like StressFX atm.
 
The game is not performing as well as it should on NVIDIA cards, mostly in regards to TressFX (which halves framerates).

Still waiting for new drivers.. NVIDIA really slipped with the 313+ releases, so many problems we've never experienced before.
 
For those wishing to compare built-in benchmark results, the custom benchmark in this article at 1680x1050 is similar to the built-in benchmark results at 1680x1050 of my system. (post 33)
 
Sorry no it isn't. It looks like those results are with the new patch which just came out so obviously we have not had a chance to re-test just yet.

Also don't forget we are testing a different part of the game which is very different to the built in benchmark.



Not so fast? You understand what "For now..." means right because that is how we started that paragraph.


Not so fast...
You obviously heard the story then... oh wait... you didn't:
"And it's worth noting that TressFX isn't the reason for Nvidia's poor showing -- that seems largely due to depth of field (DOF), which we discovered after a lengthy session of trial and error."

*tsk tsk*

So, "for now", you're telling your readers to forget about nVIDIA, and buy AMD instead? That's cool, if you don't think your readers deserve the full story.
 
Sorry no it isn't. It looks like those results are with the new patch which just came out so obviously we have not had a chance to re-test just yet.

Also don't forget we are testing a different part of the game which is very different to the built in benchmark.



Not so fast? You understand what "For now..." means right because that is how we started that paragraph.


Not so fast...
You obviously heard the story then... oh wait... you didn't:
"And it's worth noting that TressFX isn't the reason for Nvidia's poor showing -- that seems largely due to depth of field (DOF), which we discovered after a lengthy session of trial and error."

*tsk tsk*

So, "for now", you're telling your readers to forget about nVIDIA, and buy AMD instead? That's cool, if you don't think your readers deserve the full story.

Go read their previous game tests. You'll see that they recommend Nvidia a lot, I think farcry 3 was one that I recently remember. Should you expect them to not have a conclusion page?

If you go back and read the comments in these reviews, it's rather funny. When they say AMD is best you get Steve defending himself vs Nvidia fanboys and when he says Nvidia is best out comes the AMD attackers. You can see who is who.
 
That's a shame. I thought I could try this hair tech on a mainstream card. :(

I think they'll improve it. It looks very nice but is a bit overdone like the early ragdoll effect in fps games years ago(Halo comes to mine), where your guy would just bounce all over the place when you died. It's a good first attempt but they should optimize it. I'd definitely rather have it turned on. It looks better than normal by far.
 
Not so fast...
You obviously heard the story then... oh wait... you didn't:
"And it's worth noting that TressFX isn't the reason for Nvidia's poor showing -- that seems largely due to depth of field (DOF), which we discovered after a lengthy session of trial and error."

*tsk tsk*

So, "for now", you're telling your readers to forget about nVIDIA, and buy AMD instead? That's cool, if you don't think your readers deserve the full story.

I wish I knew what you were on about half the time :S
 
Surely TressFx wasn't intended to be so taxing on hardware? Hell, it really takes away the appeal when the hair just glitches and drags down performance. I remember on several occasions when playing TR my Lara would loose her hair completely lmao. A bald Lara doesn't look very sexy haha.
 
Well TressFX is still very new and it will take some time for efficiency to improve. That said the results right now are amazing, its a very cool effect. Given how much is involved in rendering hair in this kind of detail I don't imagine it was a technology designed for anything less than the most extreme GPUs.
 
Back