• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Best Graphics Card for Gaming and GPU Computing

aez5

Junior Member
I am looking for a graphics card to use both for GPU computing and for gaming on the same computer. Moreover, I'm trying not to spend more than $200. I've been looking, and it seems to me like the GTX 650 Ti is a good one. Any suggestions or thoughts?

Current system (under construction):
CPU: Core i5-3570k
MB: Gigabyte GA-Z77-DS3H
600W PSU

Thanks!
 
I am looking for a graphics card to use both for GPU computing and for gaming on the same computer. Moreover, I'm trying not to spend more than $200. I've been looking, and it seems to me like the GTX 650 Ti is a good one. Any suggestions or thoughts?

Current system (under construction):
CPU: Core i5-3570k
MB: Gigabyte GA-Z77-DS3H
600W PSU

Thanks!

http://www.newegg.com/Product/Produc...82E16814161405
walks all over the 650ti and can oc past 580 levels with a little luck
 
You'll need to be more specific about what Computing apps you'll be using. AMD is better for some and Nvidia better for others.
 
You'll need to be more specific about what Computing apps you'll be using. AMD is better for some and Nvidia better for others.

+1. Gaming also can be game specific. Even though one vendor may tend to outperform another in a given generation, if there's a specific game you really want to perform well in, it may change what recommendation is given.
 
You'll need to be more specific about what Computing apps you'll be using. AMD is better for some and Nvidia better for others.

The computing is more important than the gaming. I'll mostly be using MATLAB, but also some statistical programs like SAS, STATA, etc. My understanding is that Nvidia is better for GPU computing in general, and especially for MATLAB. Please correct me if I'm wrong.
 
Last edited:
The computing is more important than the gaming. I'll mostly be using MATLAB, but also some statistical programs like SAS, STATA, etc. My understanding is that Nvidia is better for GPU computing in general, and especially for MATLAB. Please correct me if I'm wrong.

If compute is more important than game performance don't look at the 600 series, the 500 series' Fermi architecture is much better for compute performance than Kepler.
 
If compute is more important than game performance don't look at the 600 series, the 500 series' Fermi architecture is much better for compute performance than Kepler.
I think you mean double precision compute. The 600 series is different from the 500 series. Win some, lose some.

It does come back to the type of computing. What do the programs actually do (single or double precision, heavy on memory or on number crunch, etc) and how well/at all they can use Nvidia, AMD, or both. Check the MATLAB documentation.
 
Looks pretty cut and dry as far as nVidia or AMD. MATLAB uses CUDA.
Graphics Processing Unit (GPU) for computation

To speed up computation, Parallel Computing Toolbox leverages NVIDIA GPUs that support CUDA 1.3 or later. See the full list of supported GPUs. MATLAB does not support computation acceleration using AMD/ATI GPUs at this time.

The info you need should be able to be found HERE
 
http://www.nvidia.com/object/tesla-matlab-accelerations.html

In essence you are looking for an nVidia card from the GTX 2xx series or newer. For gaming, obviously, you'd want the latest & greatest, the GTX 680. However, nVidia basically neutered their top cards this generation with regard to compute power (double precision).

Here are some MATLAB direct comparisons you should definitely find interesting.
http://blog.accelereyes.com/blog/2012/04/26/benchmarking-kepler-gtx-680/

If gaming performance takes the second priority to MATLAB you are going to be looking for a GTX 580.
 
Here are some MATLAB direct comparisons you should definitely find interesting.
http://blog.accelereyes.com/blog/2012/04/26/benchmarking-kepler-gtx-680/

If gaming performance takes the second priority to MATLAB you are going to be looking for a GTX 580.
Those *are* interesting, but unfortunately don't help much at the OP's $200 price point.

Given that the GTX660 and 560ti are about the same price, that the former has 2.5x the shaders, and that the GF114 in the older card was neutered anyway, the decision should be simple... Unless the additional memory bandwidth and dual-precision capability of a used GTX480 would make it competitive with the 660. Hard to imagine that being the case though.
 
It's $180 AR but yeah that's a nice deal too. Personally I'd rather have the MSI 560 Ti 448 is which better cooled and less expensive before rebate.
 
Thanks for the input, everyone. Sounds like the 570 or the 560 Ti are the best bets. Just for my edification, does anyone know what it is about the new 600 series (other than just the new architecture, if anything) that makes them weaker for computing?
 
Also, does the card's memory make much of a difference? I'm interested for gaming, as well as computing.
 
600 series have reduced double precision floating point units.

VRAM will help in gaming but on 1080p, the difference between 1280mb and 2048mb is not critical at all. Most of the time it'll use less than 1GB, some games push it over 1GB and only a very few (modded Skyrim for instance) would benefit from 2GB.
 
He will benefit from CUDA.

I see.

In that case I'd wait for a sale on a 660. Those 560 ti's at close to two bills are ridiculous, sorry to say.

I did the v.me sale on the PNY 660 TI for 210 bucks and that INCLUDED a copy of Assasin's Creed 3. Anyway you slice it a 560 at close to 200 is garbage.

With the emphasis on compute...I'm at LEAST getting a 570 for that price.
 
Last edited:
So in the next 10 years do you think theres gonna be another leap to something more grand in the entertainment industry of grapics an games, or are we getting closer to a maximum point.
 
Back
Top