Borderlands 2 GPU/CPU benchmarks [TechSpot/HardOCP/Others]

[railven]

<snip>

Tongue-in-cheek comment went right over your head. Haha.

Works perfectly fine for me. No slowdowns at all. Certainly not like he's reporting.

Then I have no ideas. I know our GTX 680 on a i7 930 @ 3.6ghz is not struggling as hard as my Hybrid PhysX system, and if she chugged to <20 FPS in this game I'd hear about it (as I often do - "why is my PC running so slow" gah!)

 

[wand3r3r]

How is this even conceivably possible when I'm kicking this games arse with a single GTX680 doing everything?
SLI is disabled and I choose GPU1 for PhysX and check "dedicated". eVGA monitoring tool shows 0% utilization for GPU2 and the temp stays put at idle. So I know that card is doing zip. GPU1 is chugging away.

I am owning this game with every single setting on it's highest/farthest/PhsX High/1920x1200. No matter how intense firefights get or how many characters on screen. I'm trying to understand what possible issue you could be having. Your CPU is plenty fast enough. I can only think that something is amiss with your system somehow.

How about some benchmark runs through some major boss fights? Screenshots would be helpful too. You've mentioned multiple times how you don't "notice" drops but most of us prefer some evidence. I partially believe it may be an SLI bug at this point but until evidence proves otherwise who knows.

Perhaps there is a scaling issue for SLI users or Wand3r is using an outside configuration to push his system further?

On my end, I can vouch for Keys. My GF's rig isn't as elaborate as other's and the lowest we've seen her system go to was 47 FPS. It's possible it went lower, but she hasn't mentioned it.

Hybrid PhysX isn't a cure all either, there are instances where my system has sporadoic stops as if something is loading. It would make the game unplayable, however, the GTX 680 system experiences no issues or slow downs as severe.

Either way, the PhysX effects are a great addition in this game. Sure you don't need them, but after playing the game with them I wouldn't want to play without them. Kudos nVidia on adding something of value, for once

(I don't have an outside configuration)

I don't know how it's possible but it just my experience. You can see from the other screenshots I posted they are 50-70 fps in pretty demanding scenes, but for the extreme explosion scenes it does drop.

What could be amiss? It may be an SLI bug. As I mentioned it usually is very high, but will drop on the most extreme fights. Have you guys actually benchmarked minutes of gameplay? I didn't realize that one was at 15fps until I took a screenshot, but they do noticably drop in the places with major physx.

Until I see some other screenshots with fps and MAJOR physx effects or long benchmark runs I don't have other data to prove otherwise.

I'm only showing my experience(s) and have yet to see proof of a 680 doing better then the 690 although I won't deny it's possible. A few people in the thread have admitted they get major drops but for those who claim flawless performance a little proof would be nice.

What can I use to run a benchmark for like 10 mins of action? The majority of fps will be pretty high but you will see the drops I speak of.

 

[Lonbjerg]

Tongue-in-cheek comment went right over your head. Haha.

The arguments wern't directed at you

 

[railven]

(I don't have an outside configuration)

What I meant by that, was if you were pushing higher AA through Inspector.

I don't know how it's possible but it just my experience. You can see from the other screenshots I posted they are 50-70 fps in pretty demanding scenes, but for the extreme explosion scenes it does drop.

What could be amiss? It may be an SLI bug. As I mentioned it usually is very high, but will drop on the most extreme fights. Have you guys actually benchmarked minutes of gameplay? I didn't realize that one was at 15fps until I took a screenshot, but they do noticably drop in the places with major physx.

Until I see some other screenshots with fps and MAJOR physx effects or long benchmark runs I don't have other data to prove otherwise.

I'm only showing my experience(s) and have yet to see proof of a 680 doing better then the 690 although I won't deny it's possible. A few people in the thread have admitted they get major drops but for those who claim flawless performance a little proof would be nice.

I can post a video of GTX 680 vs Radeon 7970+GTX 460, but I don't have video editing software. So it would just be two youtube clips haha. I'll try to compile something tonight after work.

Not sure what could cause your dips, and perhaps it can be related to my dips, but I know my Hybrid system can trip even as into the 40's, while the GF's GTX 680 (same environment since we play Co-OP together) doesn't experience a hitch (or at least when I ask "whats your FPS" she responds with "60, no wait 58" then I look at mine, "man, I'm getting 44".)

That usually is followed by a laugh since she loves to 1uP me haha.

 

[Jaydip]

In case anyone interested
http://www.eurogamer.net/articles/digitalfoundry-borderlands-2-face-off

 

[SirPauly]

Zogrim offers his views on Borderlands 2:

GPU PhysX in Borderlands 2

http://physxinfo.com/news/9500/gpu-physx-in-borderlands-2/

 

[Termie]

In case anyone interested
http://www.eurogamer.net/articles/digitalfoundry-borderlands-2-face-off

Awesome article, thanks! I'll add it to the OP.

 

[Keysplayr]

Zogrim offers his views on Borderlands 2:

GPU PhysX in Borderlands 2

http://physxinfo.com/news/9500/gpu-physx-in-borderlands-2/

Awesome awesome PhysX comparison there. He does some nice videos.
BL2 really pours on the PhysX. I especially like how the pile drivers send vibrations through the air and the ground and force gravel and dirt away like a shockwave. That is a first.

 

[Keysplayr]

How about some benchmark runs through some major boss fights? Screenshots would be helpful too. You've mentioned multiple times how you don't "notice" drops but most of us prefer some evidence. I partially believe it may be an SLI bug at this point but until evidence proves otherwise who knows.

I can do all of that, but it's kind of a waste of time proving to you that what I said I am getting I am truly getting. I'd much rather experiment with your system to try and find out what the issue is, and believe me, there is an issue.
Are you up to it?

 

[Termie]

Zogrim offers his views on Borderlands 2:

GPU PhysX in Borderlands 2

http://physxinfo.com/news/9500/gpu-physx-in-borderlands-2/

Awesome awesome PhysX comparison there. He does some nice videos.
BL2 really pours on the PhysX. I especially like how the pile drivers send vibrations through the air and the ground and force gravel and dirt away like a shockwave. That is a first.

Added to OP.

 

[Grooveriding]

How about some benchmark runs through some major boss fights? Screenshots would be helpful too. You've mentioned multiple times how you don't "notice" drops but most of us prefer some evidence. I partially believe it may be an SLI bug at this point but until evidence proves otherwise who knows.


(I don't have an outside configuration)

I don't know how it's possible but it just my experience. You can see from the other screenshots I posted they are 50-70 fps in pretty demanding scenes, but for the extreme explosion scenes it does drop.

What could be amiss? It may be an SLI bug. As I mentioned it usually is very high, but will drop on the most extreme fights. Have you guys actually benchmarked minutes of gameplay? I didn't realize that one was at 15fps until I took a screenshot, but they do noticably drop in the places with major physx.

Until I see some other screenshots with fps and MAJOR physx effects or long benchmark runs I don't have other data to prove otherwise.

I'm only showing my experience(s) and have yet to see proof of a 680 doing better then the 690 although I won't deny it's possible. A few people in the thread have admitted they get major drops but for those who claim flawless performance a little proof would be nice.

What can I use to run a benchmark for like 10 mins of action? The majority of fps will be pretty high but you will see the drops I speak of.

You're not imagining it. I get slowdowns out of nowhere as well with low gpu usage. Sometimes it is just from a certain area or when viewing a certain landscape. Other times it is when heavy physx effects are going off everwhere. That part is pretty consistent with gpu physx titles in the past. Gpu physx is a performance killer when used heavily, as it tends to be on high settings.

Nvidia has driver issues in Borderlands 2 as well.


http://www.neogaf.com/forum/showpost.php?p=42326717&postcount=708

So that is another problem. In multiplayer with the excessive physx debris flying around, I get nasty drops to 20-30fps. My CPU is no slouch as well, so it is definitely GPU/driver/physx related. Nvidia needs to get some fixed drivers out for this game.

 

[wand3r3r]

I can do all of that, but it's kind of a waste of time proving to you that what I said I am getting I am truly getting. I'd much rather experiment with your system to try and find out what the issue is, and believe me, there is an issue.
Are you up to it?

What does it entail? I can try figure out what's going on if you have some new ideas. (when I get a chance)

That one dude a couple pages back was claiming "I don't have the datasets" motivated me to try one GPU as dedicated. I'm not concerned with the datasets, I am concerned with my experience.

So now I was playing for a little while with one other person using 1 gpu for graphics and the other for dedicated physx. I can see that there is a huge difference between battles with 2 people versus the 4 where I had previously seen massive slowdowns so I can hardly compare the runs. It played fine (always does) but I forgot to take screenshots after the first few minutes and the lowest I recorded even with some not to major fights was 39 fps. With major fights I'm sure it would have dropped much lower with four players (the enemies and effects get crazy with 4 players and physx ).

Nvidia has driver issues in Borderlands 2 as well.

http://www.neogaf.com/forum/showpost.php?p=42326717&postcount=708

LOL at the link.

 

[SirPauly]

Zogrim touched on this:

Of course, heavy scenes with lots of fluids and particles, especially when you are playing in co-op mode, will result in massive slowdowns, but still – it is a good progress.

 

[SirPauly]

Awesome awesome PhysX comparison there. He does some nice videos.
BL2 really pours on the PhysX. I especially like how the pile drivers send vibrations through the air and the ground and force gravel and dirt away like a shockwave. That is a first.

He also has an artsy flair with his contributions with the choice of music and how it flows with the video.

Really enjoy the PhysX force field effects and turbulence effects.

 

[WMD]

Really nice implementation of Physx in borderlands 2. This is one of the things that really help differentiate GeForce cards from Radeons. That said, IMO Physx is still far behind Havok in terms of performance.

http://www.youtube.com/watch?nomobile=1&v=wJUWFz8JP88

All running on the cpu no slowdowns. The particle effects are almost as good as those in borderlands.

 

[Lonbjerg]

Really nice implementation of Physx in borderlands 2. This is one of the things that really help differentiate GeForce cards from Radeons. That said, IMO Physx is still far behind Havok in terms of performance.

http://www.youtube.com/watch?nomobile=1&v=wJUWFz8JP88

All running on the cpu no slowdowns. The particle effects are almost as good as those in borderlands.

Are you serious?

 

[Keysplayr]

Just played BL2 multi. Everybody is familiar with the army of robots battle? We were eight fighters against them. There was so much (expletive) flying around I almost didn't know which way was up. The lowest my fps went was 47 fps. I kept a close eye on it. Again, this is with all settings highest/farthest/PhysX High 1920x1200. Is there a specific area in this game I should be investigating? Because no matter how I try, I can't find any areas that make my fps tank. Not a one. I could be always missing it, so tell me where to go.

Wand3r3r, Grooveriding? Where to?

 

[f1sherman]

Are you serious?

Well on the face side they might look similar.
But Havok is all about ragdoll and collision. While it's destruction model is hardly a physics and mostly just a set of scripted events.

PhysX otoh really tries to be a real physical model. Something like this sound wise.
For example you don't have this in any other physics model:

Each of these effects is generated by manipulating tens of thousands of particles that persist in a realistic manner using physics calculations, ensuring they don&#8217;t clip through terrain and are affected by other forces, such as explosions.
&#8211; if an effect clips through a wall you&#8217;re brought out of the action and reminded that it&#8217;s an illusion. This realism also gives hardware PhysX an advantage over software solutions, which use approximated effects, meaning they lack the realistic properties of PhysX&#8217;s particles. In other words, software smoke clips through walls; hardware-accelerated PhysX smoke hits the wall and realistically floats up towards the sky.

You just don't have readily available Clothing system, fluids and turbulence in any other physics model.

 

[Lonbjerg]

Well on the face side they might look similar.
But Havok is all about ragdoll and collision. While it's destruction model is hardly a physics and mostly just a set of scripted events.

PhysX otoh really tries to be a real physical model. Something like this sound wise.
For example you don't have this in any other physics model:

Each of these effects is generated by manipulating tens of thousands of particles that persist in a realistic manner using physics calculations, ensuring they don&#8217;t clip through terrain and are affected by other forces, such as explosions.
&#8211; if an effect clips through a wall you&#8217;re brought out of the action and reminded that it&#8217;s an illusion. This realism also gives hardware PhysX an advantage over software solutions, which use approximated effects, meaning they lack the realistic properties of PhysX&#8217;s particles. In other words, software smoke clips through walls; hardware-accelerated PhysX smoke hits the wall and realistically floats up towards the sky.

You just don't have readily available Clothing system, fluids and turbulence in any other physics model.

The reason I asked him if he was serious was because he is either blind...or a troll.

*sign*, Lonbjerg, what are we supposed to do with you?
-ViRGE

 

[Keysplayr]

Nm

 

[boxleitnerb]

Can someone try an explanation why forcing MSAA/SGSSAA works at some resolutions, but not at others?

Works with:

1920x1080
2560x1440
2560x1600
3840x2160

Doesn't work with:

1600x900
1680x1050
2880x1620

This was already seen with Mirror's Edge, which is Unreal Engine, too.

 

[f1sherman]

can someone explain why SGSSAA works without MSAA working

has to do with with MSAA being hidden, but still existent because of planned TXAA.

 

[boxleitnerb]

Yeah it's weird - often you can force SGSSAA with special compatibility bits, but those bits don't work with plain MSAA. But you're right, the basis for SGSSAA is still MSAA, however that works in these cases.

 

[WMD]

Well on the face side they might look similar.
But Havok is all about ragdoll and collision. While it's destruction model is hardly a physics and mostly just a set of scripted events.

PhysX otoh really tries to be a real physical model. Something like this sound wise.
For example you don't have this in any other physics model:

Each of these effects is generated by manipulating tens of thousands of particles that persist in a realistic manner using physics calculations, ensuring they don’t clip through terrain and are affected by other forces, such as explosions.
– if an effect clips through a wall you’re brought out of the action and reminded that it’s an illusion. This realism also gives hardware PhysX an advantage over software solutions, which use approximated effects, meaning they lack the realistic properties of PhysX’s particles. In other words, software smoke clips through walls; hardware-accelerated PhysX smoke hits the wall and realistically floats up towards the sky.

You just don't have readily available Clothing system, fluids and turbulence in any other physics model.

Not true. Havok is not scripted and its definitely capable of simulating cloth physics.

http://www.youtube.com/watch?v=8qRcp2zFdxg

 

[WMD]

Are you serious?

Are you serious? You are either blind or a troll.