you would think a gtx570 could max any game at 1280...

[WMD]

1280x960

6950 84.5
5850 80.3

Difference 5.23%

1680x1050

6950 69.6
5850 64.2
Difference = 8.41%

Can you tell me the memory clock of your card?? mine is at default 1250MHz.

Memory bandwidth may play a bigger role than tess in higher ress in this bench.

I am running at 810 core 1250memory. VLIW4 has more TMUs.

 

[tweakboy]

You should get 300fps steady without a drop at that weak crummy resolution of 1280x1024...


Try

1080p with 8xAA and 16xAF vsync on,high quality and Im stuck on 60fps on everygame,, eg.. Crysis 2 , Mass Effect 2 , Dirt 3 etc.....

Soo sorry but your OS your hardware is messed up. Lets help you fix that.

 

[toyota]

You should get 300fps steady without a drop at that weak crummy resolution of 1280x1024...


Try

1080p with 8xAA and 16xAF vsync on,high quality and Im stuck on 60fps on everygame,, eg.. Crysis 2 , Mass Effect 2 , Dirt 3 etc.....

Soo sorry but your OS your hardware is messed up. Lets help you fix that.

please go away

 

[RussianSensation]

I don't think these two are quite so equal in newer games and console ports.

http://www.tomshardware.com/charts/...re,746.html?prod%5B2143]=on&prod[2093]=on

The 7900 is no match for the X1900 in console ports Call of Duty 4, Mass Effect, and Asassin's Creed. The X1900 is almost twice as fast in CoD.

The X1900 is also appreciably faster in Crysis and HL2: E2, but perhaps not by an abnormal amount.

In the non-console-ports (FS, WiC, ETQW) they either tie or the 7900 is a bit faster.

Ya, you are right!! I was a bit too generous towards the 7900GTX since at the time it was close to the X1900XT series. It seems the X1900XT turned out to be better in the long term over the 7900GTX. Even the 8600GTS spanked 7900GTX when games got more modern/shader heavy. X1900XT is still a slower than a 8800GTS 320mb and a GTX570 is likely still about 4-5x faster than that with AA and advanced DX11 features on. My general point was that the DX11 code is slower in Metro 2033 vs. DX10 to begin with and once you go from High to Very High in IQ, the game gets very demanding vs. the relative graphical benefit VH vs. H brings.

Even without DOF setting, at VHQ settings with 4x MSAA, this game is killer:

High - GTX570 = 40 fps

Very High - GTX570 = 27 fps (ouch!)

Source

Based on some modern games that GameGPU reviewed, it seems current generation of cards (i.e., GTX470/480/570/580) provide about the same performance in DX11 as previous generation (i.e., GTX260 216/275/280/285) did in DX9/10. So you get 'free' DX11 in a lot of games. Metro 2033 though is more brutal than Crysis. I think it'll need a GTX670 SLI or faster to max it out at 60 fps at all times with DOF and all.

 

[toyota]

maybe DX11 is not any slower in this game. I just ran the bench using the same settings they did and then in DX10. I got the same framerate in both cases even though DX10 does not have tessellation.


Options: Resolution: 1280 x 960; DirectX: DirectX 11; Quality: Very High; Antialiasing: MSAA 4X; Texture filtering: AF 16X; Advanced PhysX: Disabled; Tesselation: Enabled; DOF: Disabled

•Total Frames: 2766, Total Time: 59.81432 sec
•Average Framerate: 46.34
•Max. Framerate: 106.20 (Frame: 2475)
•Min. Framerate: 6.68 (Frame: 6)


Options: Resolution: 1280 x 960; DirectX: DirectX 10; Quality: Very High; Antialiasing: MSAA 4X; Texture filtering: AF 16X; Advanced PhysX: Disabled; Tesselation: Not supported; DOF: Not supported

•Total Frames: 2806, Total Time: 59.86141 sec
•Average Framerate: 46.98
•Max. Framerate: 107.75 (Frame: 3)
•Min. Framerate: 6.24 (Frame: 6)

 

[AtenRa]

Ok i re-run the AvP bench with Tessellation On/Off to see what happens, same settings as before same hardware.

First run with Tess on, second run with Hardware Tess Off in the benchmark and third run with Tess level to Zero in the AMD CCC panel.

It seams that Tess doesn't stress the card that much.

I have found a video showing Tess On/Off, seams they use very high tess levels on the Alien/Predator models with dynamic Tessellation.

AvP PC Game Wireframe Showing Tessellation

Edit: Does anyone knows if the AvP bench uses high level of Tess or not ??

 

[WMD]

maybe DX11 is not any slower in this game. I just ran the bench using the same settings they did and then in DX10. I got the same framerate in both cases even though DX10 does not have tessellation.


Options: Resolution: 1280 x 960; DirectX: DirectX 11; Quality: Very High; Antialiasing: MSAA 4X; Texture filtering: AF 16X; Advanced PhysX: Disabled; Tesselation: Enabled; DOF: Disabled

•Total Frames: 2766, Total Time: 59.81432 sec
•Average Framerate: 46.34
•Max. Framerate: 106.20 (Frame: 2475)
•Min. Framerate: 6.68 (Frame: 6)


Options: Resolution: 1280 x 960; DirectX: DirectX 10; Quality: Very High; Antialiasing: MSAA 4X; Texture filtering: AF 16X; Advanced PhysX: Disabled; Tesselation: Not supported; DOF: Not supported

•Total Frames: 2806, Total Time: 59.86141 sec
•Average Framerate: 46.98
•Max. Framerate: 107.75 (Frame: 3)
•Min. Framerate: 6.24 (Frame: 6)

DX10 and DX11 performs exactly the same on my system.

DX9 and DX11 comparison. Very high 1680x1050. AA, DOF, Tessellation off

Can you tell which is which?

 

[Via]

Stills don't tell the whole picture though.

Those effects are necessary imo to get the full effect of the game.

 

[apoppin]

DX10 and DX11 performs exactly the same on my system.

DX9 and DX11 comparison. Very high 1680x1050. AA, DOF, Tessellation off

Can you tell which is which?

Your Fraps counter says different

Any if you have DX11, turn Tessellation On, so it is easy to see the difference in a still shot. Why would you run with it Off, unless your system is too slow?

 

[toyota]

Your Fraps counter says different

Any if you have DX11, turn Tessellation On, so it is easy to see the difference in a still shot. Why would you run with it Off, unless your system is too slow?

his screenshots are a "DX9 and DX11 comparison"

 

[tviceman]

his screenshots are a "DX9 and DX11 comparison"

I think the point apoppin is making is that trying to spot a difference between dx9 and dx11 is pointless when the only difference in the screenshots is the rendering pipeline. If the screenshots aren't using dx11 features, such as the DOF and tessellation then there should be very, very little (if any) difference in the comparison shots.

 

[apoppin]

his screenshots are a "DX9 and DX11 comparison"

He mentions no performance difference between DX10 and DX 11; without AA, DOF or Tessellation (obviously). i said to turn Tessellation on and then you would see the difference in all of his screenshots (as well as seeing the performance difference between DX10 and DX11).

And i mentioned the Fraps framerate as the clue which shot was taken on which DX pathway; it is easy (on his shots) to tell accurately which is DX9 and which is DX11.
:whiste:

 

[toyota]

I think the point apoppin is making is that trying to spot a difference between dx9 and dx11 is pointless when the only difference in the screenshots is the rendering pipeline. If the screenshots aren't using dx11 features, such as the DOF and tessellation then there should be very, very little (if any) difference in the comparison shots.

I do not see what that has to do with the comment. he clearly looked at "DX10 and DX11 performs exactly the same on my system" and he then said "Your Fraps counter says different". it looks like he simply thought the screens were DX10 and DX11.

 

[RussianSensation]

Your Fraps counter says different

Any if you have DX11, turn Tessellation On, so it is easy to see the difference in a still shot. Why would you run with it Off, unless your system is too slow?

His entire point was to show that DX11 code runs slower than DX9 with no difference in visuals, unless you enable the 3 killer features in this game: Adaptive Anti-Aliasing, DOF and Tessellation.

Tessellation barely makes a difference in this particular game. That leaves AAA and DOF as the 2 key features. DOF is brutal on modern hardware.

 

[apoppin]

I do not see what that has to do with the comment.

It sounds like a personal issue to me. And perhaps your system is a bit slow at a low resolution just like your last two CPUs that you complained about and no one else could replicate.
:whiste:

i have to get back to work. Just checking in. Aloha


This is a thread derail and is not acceptable. The thread has nothing to do with toyota's CPUs and his experience with them. Bringing them up this way only starts a derail, and a personal one at that.

Moderator jvroig

 

[toyota]

It sounds like a personal issue to me. And perhaps your system is a bit slow at a low resolution just like your last two CPUs that you complained about and no one else could replicate.
:whiste:

i have to get back to work. Just checking in. Aloha

personal issue with what? I am just commenting that it looks like you thought the screens were DX10/DX11 comparisons. why do you always have to act like an [redacted] you are no better than AzN that was banned.

No profanity please. Leave the complaints to the reported post function.

Moderator jvroig

 

[apoppin]

personal issue with what? I am just commenting that it looks like you thought the screens were DX10/DX11 comparisons. why do you always have to act like an ass and try to stir up old shit? you are no better than AzN that was banned.

A personal issue with the way you read and misinterpret my posts

Perhaps it's relevant to bring up your consistently low framerates with your CPUs. You are the one that posts odd things about CPUs; including recently posting in CPU forum that you felt the Phenom II "felt snappier" than your C2D on the desktop. Your observations seem to be unique to you and your PCs.


You are also the one that tried to nitpick my comment even after i completely explained it.


i think if you have an issue with me personally, you should take it to PMs without attempting to derail your own thread.

This is a thread derail and is not acceptable. The thread has nothing to do with toyota's CPUs and his experience with them. Bringing them up this way only starts a derail, and a personal one at that.

And re: "you should take it to PMs without attempting to derail your own thread" - you are the one who started the derail and made it personal. Stop doing that. If you feel the need to bring up toyota's CPUs and how he is completely wrong about his experience, make your own thread or keep them to yourself.

Moderator jvroig

 

[toyota]

A personal issue with the way you read and misinterpret my posts

Perhaps it's relevant to bring up your consistently low framerates with your CPUs. You are the one that posts odd things about CPUs; including recently posting in CPU forum that you felt the Phenom II "felt snappier" than your C2D on the desktop. Your observations seem to be unique to you and your PCs.


You are also the one that tried to nitpick my comment even after i completely explained it.


i think if you have an issue with me personally, you should take it to PMs without attempting to derail your own thread.

hey genius, look at the time I posted my comment. sure it came after your better explanation but I had already submitted before i saw that.

and do not start [redacted] in a thread by trying to use irrelevant past topics and then tell me to take it to PM.

Agreed on your last point, but still no profanity please.

Moderator jvroig

 

[apoppin]

Irrelevant? i think you are very consistently getting low fps with every recent CPU you have had.

. . . and take the personal stuff to PMs
.. genius, yourself

 

[WMD]

Your Fraps counter says different

Any if you have DX11, turn Tessellation On, so it is easy to see the difference in a still shot. Why would you run with it Off, unless your system is too slow?

My system is too slow to even run the above screenshot settings. The fps will drop to 30s in areas with strong volumetric lighting making aiming very difficult. I usually play on high settings.

His entire point was to show that DX11 code runs slower than DX9 with no difference in visuals, unless you enable the 3 killer features in this game: Adaptive Anti-Aliasing, DOF and Tessellation.

Tessellation barely makes a difference in this particular game. That leaves AAA and DOF as the 2 key features. DOF is brutal on modern hardware.

You are right. The tessellation is very subtle in this game so its best left off even if its only 10% hit on fps. AAA is enabled by default in very high settings. MSAA and DOF are the performance killers.

It is common knowledge that DX11 code-path is more efficient than DX9 and will perform faster when DX11 effects are switched off. But here it performs slower so I am really not sure if there is any image quality difference that I overlooked.

 

[toyota]

DX9 disables AAA but I cannot even tell.

 

[Via]

My system is too slow to even run the above screenshot settings. The fps will drop to 30s in areas with strong volumetric lighting making aiming very difficult. I usually play on high settings.




You are right. The tessellation is very subtle in this game so its best left off even if its only 10% hit on fps. AAA is enabled by default in very high settings. MSAA and DOF are the performance killers.

It is common knowledge that DX11 code-path is more efficient than DX9 and will perform faster when DX11 effects are switched off. But here it performs slower so I am really not sure if there is any image quality difference that I overlooked.

I've been messing with every setting trying to wring as many fps as possible for eyefinity, and there's a pretty big difference to my eyes. I didn't like the visuals as much with the advanced features off.

 

[RussianSensation]

It sounds like a personal issue to me. And perhaps your system is a bit slow at a low resolution just like your last two CPUs that you complained about and no one else could replicate.
:whiste:

i have to get back to work. Just checking in. Aloha

Many websites such as PCgameshardware, LegionHardware, Techspot, Gamegpu.ru, Xbitlabs reported that dual-core CPUs were limiting GPUs in the last 2+ years. Not everyone agreed with this assessment. I don't think it's accurate to say that no one else could replicate slower performance with dual core CPUs in Resident Evil 5, GTAIV, Supreme Commander 2, Far Cry 2, Arma 2/3, BF:BC2, etc.

toyota was one of the few guys on our forum out of 100s of users who would bring up specific games such as BF:BC2 where his dual core E8500/E8600 (forget which one) was often not giving him optimal performance. He did this so that other enthusiasts wouldn't throw hundreds of dollars on GTX570/580/6970 hardware and then be disappointed since they wouldn't be getting optimal performance due to their slower CPUs.

toyota also provided feedback to users who had similar GPUs as him (i.e., 4870/4890/GTX260/275, etc.) that upgrading to a much faster GPU (GTX470 in his case) wouldn't solve their problems in many modern games due to CPU limitations. He did this despite many users who claimed that his performance was somehow unusual, somehow related to his "poor setup: drivers, bios, etc. etc.". At least 4-5 professional review websites also showed similar findings to toyota's that C2D generation (or Athlon II CPUs for that matter), even when overclocked, was simply bottlenecking many modern GPUs.

Even now, with more and more games coming out that use more than 2 threads, we still have threads where people ignore CPU limitations and purchase $250-300 GPUs to pair with their stock E6800/E8400/Q6600/Q9550 CPUs.

 

[Grooveriding]

Many websites such as PCgameshardware, LegionHardware, Techspot, Gamegpu.ru, Xbitlabs reported that dual-core CPUs were limiting GPUs in the last 2+ years. Not everyone agreed with this assessment. I don't think it's accurate to say that no one else could replicate slower performance with dual core CPUs in Resident Evil 5, GTAIV, Supreme Commander 2, Far Cry 2, Arma 2/3, BF:BC2, etc.

toyota was one of the few guys on our forum out of 100s of users who would bring up specific games such as BF:BC2 where his dual core E8500/E8600 (forget which one) was often not giving him optimal performance. He did this so that other enthusiasts wouldn't throw hundreds of dollars on GTX570/580/6970 hardware and then be disappointed since they wouldn't be getting optimal performance due to their slower CPUs.

toyota also provided feedback to users who had similar GPUs as him (i.e., 4870/4890/GTX260/275, etc.) that upgrading to a much faster GPU (GTX470 in his case) wouldn't solve their problems in many modern games due to CPU limitations. He did this despite many users who claimed that his performance was somehow unusual, somehow related to his "poor setup: drivers, bios, etc. etc.". At least 4-5 professional review websites also showed similar findings to toyota's that C2D generation (or Athlon II CPUs for that matter), even when overclocked, was simply bottlenecking many modern GPUs.

Even now, with more and more games coming out that use more than 2 threads, we still have threads where people ignore CPU limitations and purchase $250-300 GPUs to pair with their stock E6800/E8400/Q6600/Q9550 CPUs.

Man those RO2 benchmarks are abysmal. Such an unoptimized POS, and that is the new unreal engine under DX9. I get about 45-70FPS depending on the scene in this game.

 

[WMD]

DX9 disables AAA but I cannot even tell.

AAA is analytical anti aliasing and enabled in both Dx9 and Dx11 very high quality mode.