MSI 560ti 448 Performance issue

[fluffywormwood]

Here is a link to the card I just installed

http://www.newegg.com/Product/Product.aspx?Item=N82E16814127615

I just installed this card last night and i'm not to happy with its performace. I tried using the Furmark benchmark and I have everything in the Nvidia panel geared to performance and during the benchmark my fps starts off at 25 then drops to 21 then to 23 and it's very very choppy. Now if i go back into the gettings and click the Quality preset. I get 25 fps and its smooth.

Now 2 days before I had a PNY 560ti and at stock levels with no OC i was getting aroud 27-28 FPS on the performance setting and with an OC I was up to 31 FPS. A week prior to this I had the PNY 550ti and I managed around 18-20 FPS. The 550 & 560 were both stable with no issues.

I installed the driver from the disk that came with the card testing it and then downloaded the new beta drivers from Nvidia and the results are the same, so I don't think there is a driver issue. MSI customer support didn't provide much help either.

Here is the system that I have it installed in

Dell XPS 8100
650 i5 3.2ghz
8GB of ram
PC Power & Cooling 600 psu
1TB HD

The first thing that did come to my mind this morning was is it at all possible there is a bottleneck somewhere? The 560 448 isnt THAT much better than a 560 and the 560 was fine so I don't know if my system cant handle those additional cores?

I have plenty of power

Any insight for me?

 

[weezys_advocate]

in the last week you have upgraded from a 550ti to a 560ti to a 560 ti 448 ? why .... also did you do a clean driver instal ? if not please do so with latest WHQL... also when you are benchmarking is your GPU usage 95-100% ?

 

[fluffywormwood]

Yeah I know..I have issues..

I uninsalled every driver I had then installed from the disk so if thats what you call a clean install.

Yeah when im in the bench marks the gpu usage is between 95-100% (is that where it should be)?

 

[Novuake]

OK hope I can explain this adeqeutly... furmark is designed to push your graphics as far as possible to get the best possible idea of the ceiling of the card. If it is able to handle more intense graphics rendering Furmark will increase intencity so that you always end up with under 30fps.

The score at the end of the bench is the way to compare, not the fps or the choppyness in the test itself.

 

[weezys_advocate]

yes it should be between 95-100% this means there is no bottleneck in your system . i would also suggest restoring defualts on the nvidia pannel.... good spot by Novuake too ^^

 

[MKBL]

As Novuake mentioned, it seems that Futuremark tests show very low fps no matter what card it is run on. My former Gigabyte GTX 460 and the current EVGA GTX 560 Ti 448 Classified showed similar fps, but eventually the latter had much higher score.

 

[inzone]

If you want real world performance results then get a program like Fraps and play a game like BF3 or MW3 and see what the game fps is and you can judge the cards performance from that. To do a benchmark with Furmark is not the same as playing the game which is what you are getting the card for. You don't buy a card so you can sit there and run benchmarks. Have you played any games with the card yet?

 

[fluffywormwood]

OK hope I can explain this adeqeutly... furmark is designed to push your graphics as far as possible to get the best possible idea of the ceiling of the card. If it is able to handle more intense graphics rendering Furmark will increase intencity so that you always end up with under 30fps.

The score at the end of the bench is the way to compare, not the fps or the choppyness in the test itself.

Ok I feel much better than. Ill try the benchmarks again when I get home tonight.

The thing that's throwing me off is the 560 and 550 were smooth and not choppy and I just assumed the 448 would perform better than the 560 by showing a higher fps during the benchmark.

 

[MKBL]

You might have run the benchmark with 560 and 550 in lower setting, then. Did you run them on exactly same setting and the same version of 3DMark?

 

[fluffywormwood]

You might have run the benchmark with 560 and 550 in lower setting, then. Did you run them on exactly same setting and the same version of 3DMark?

same version and same settings.

 

[daveface]

I would suggest uninstalling your drivers from Add/Remove Programs, reboot into safemode "F8 at boot menu" and run Driver Cleaner to delete the registry values left behind, and then reinstall the newest driver. Nvidia has released a new beta driver that should be newer than the one that came on the CD. I would see if this new driver is compatible with your card... if it isn't, re-install the drivers from the CD. Also, these cards have some headroom in the OC http://www.guru3d.com/article/geforce-gtx-560-ti-448-core-review/23

 

[DelroyMonjo]

Downoad Unigine Heaven 2.5 and run it at whatever it default to.
It's repeatable and only takes a few minutes to run.

 

[Novuake]

OMW can you people stop giving him advice that he does not need! Its a benchmark and he is not looking at the score! Geez...

 

[monsta]

Why do people always refer to furmark as a real time perfromance benchmark?
There is probably nothing worng with your card.
Play a game use fraps, compare fps and and then compare the difference, furmark is to push the boundaries of your card , its more a of testing tool to see if you card can handle being thrashed.
If you really have to, 3dmark is also good to use to compare your systems performance with different gpus and also to see if your score is as good as others with a similar system to you.

 

[AMD Radeon]

hey you use wrong benchmark! furmark is for testing wheter the card has artifact/defect or not. to nech FPS use dirt 3 game you will see the difference

 

[fluffywormwood]

Powered by Unigine Engine

Heaven Benchmark v2.5 Basic

FPS:
30.6
Scores:
772
Min FPS:
10.5
Max FPS:
75.6
Hardware

Binary:
Windows 32bit Visual C++ 1600 Release Mar 1 2011
Operating system:
Windows 7 (build 7601, Service Pack 1) 64bit
CPU model:
Intel(R) Core(TM) i5 CPU 650 @ 3.20GHz
CPU flags:
3191MHz MMX SSE SSE2 SSE3 SSSE3 SSE41 SSE42 HTT
GPU model:
NVIDIA GeForce GTX 560 Ti 8.17.12.9053 1280Mb
Settings

Render:
direct3d11
Mode:
1920x1080 8xAA fullscreen
Shaders:
high
Textures:
high
Filter:
trilinear
Anisotropy:
16x
Occlusion:
enabled
Refraction:
enabled
Volumetric:
enabled
Tessellation:
normal

All settings are MAXED

I tried furmark again and my scores are the same

It's very difficult for me to not compare benchmarks unless someone here can explain to why is it that a 550 and 560 can score higher than a 448 . The conditions are all the same except for the gpu


http://www.ozone3d.net/benchmarks/furmark_192_score.php?id=25960

above is a link to my most recent score

 

[fluffywormwood]

also for some reason it stills thinks i have a 560ti and not a 448

 

[MKBL]

Do you have integrated graphic enabled? When I enabled mine, some software read a few information wrong, i.e. GTX 560 Ti 448 as just 560 Ti, and HD 2000 as HD 1000. I didn't run a benchmark in this condition, though.

My system:

i5 2400
MSI H67MAE45 B3
EVGA GTX 560 Ti 448

 

[fluffywormwood]

hmm where is the intergrated graphics option

 

[nforce4max]

If I had your rig I would have upgraded the cpu instead. The i5 650 isn't a true quad and the only thing that games well these days is just that a true quad core. Beyond that any thing more doesn't provide much of a bang for the cost and below that doesn't have the power to really handle everything. I know that when I bought a i5 760 I knew that I was splurging on something that wouldn't get much use but that is what you might look into or a i7 875k if you can ever find one at a low price. Why the 875k well its unlocked for this platform and then you can software overclock on that dell.

 

[MKBL]

In BIOS of my system, I can choose which graphics should be priority on boot, and also there is an option to enable/disable dual display setup. When I want to use two monitors, I set the booting graphic to be PCI, which is the GTX card, and enable integrated graphics for dual display.

 

[fluffywormwood]

If I had your rig I would have upgraded the cpu instead. The i5 650 isn't a true quad and the only thing that games well these days is just that a true quad core. Beyond that any thing more doesn't provide much of a bang for the cost and below that doesn't have the power to really handle everything. I know that when I bought a i5 760 I knew that I was splurging on something that wouldn't get much use but that is what you might look into or a i7 875k if you can ever find one at a low price. Why the 875k well its unlocked for this platform and then you can software overclock on that dell.

This rig came with a GT 220. I figured that was the weakest link next to my PSU.

 

[inzone]

also for some reason it stills thinks i have a 560ti and not a 448

The 448 is a 560Ti , if you look at the description of the card it is an EVGA 012-P3-2066-KR GeForce GTX 560 Ti (Fermi) 448 Cores FTW 1280MB 320-bit GDDR5 PCI Express 2.0 x16 HDCP Ready SLI Support Video Card.