Overclock.net › Forums › Graphics Cards › NVIDIA › GPUs and Matlab
New Posts  All Forums:Forum Nav:

GPUs and Matlab

post #1 of 7
Thread Starter 
I have need to do parallel processing in Matlab and am trying to disentangle what hardware is compatible and would be useful.

Matlab's author says I need a "NVIDIA CUDA-enabled device with compute capability of 1.3 or greater"
But Nvidia does not supply the compute capability of many of their GPUs.
However, Wikipedia claims to know them
Does anyone have experience or insight in this regard?
    
CPUMotherboardGraphicsRAM
i7 950 ASUS Sabertooth GeForce 9500 GT 12GB Corsair Dominator 
Hard DriveOSMonitorPower
Cruical 64GB SSD / Seagate 7200 1TB Windows 7 Enterprise (2) Viewsonic 23.5" HDMI Corsair 850TX 
Case
HAF 932 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
i7 950 ASUS Sabertooth GeForce 9500 GT 12GB Corsair Dominator 
Hard DriveOSMonitorPower
Cruical 64GB SSD / Seagate 7200 1TB Windows 7 Enterprise (2) Viewsonic 23.5" HDMI Corsair 850TX 
Case
HAF 932 
  hide details  
Reply
post #2 of 7
Open GPU-Z and hover your mouse over the checkmark next to CUDA and it will tell you the supported compute capability.

EDIT: Pretty sure that 1.3 is GT200 if memory serves right.
Edited by malayalam - 3/3/11 at 3:54pm
Walrus
(13 items)
 
  
CPUMotherboardGraphicsRAM
Q6600 Asus P5Q-E Gigabyte GTX 470 SOC  Kingston 4 GB (Cheap and generic...) 
Hard DriveOSMonitorPower
Crucial M4 64 GB + Seagate 160GB 7200rpm + WD 1... Windows 7 x64 Pro / Xubuntu 11.10 x 64 ASUS VH226H 21.5" 2ms 1920x1080 Corsair TX750 
Case
Cooler Master 922 
  hide details  
Reply
Walrus
(13 items)
 
  
CPUMotherboardGraphicsRAM
Q6600 Asus P5Q-E Gigabyte GTX 470 SOC  Kingston 4 GB (Cheap and generic...) 
Hard DriveOSMonitorPower
Crucial M4 64 GB + Seagate 160GB 7200rpm + WD 1... Windows 7 x64 Pro / Xubuntu 11.10 x 64 ASUS VH226H 21.5" 2ms 1920x1080 Corsair TX750 
Case
Cooler Master 922 
  hide details  
Reply
post #3 of 7
Thread Starter 
Quote:
Open GPU-Z and hover your mouse over the checkmark next to CUDA and it will tell you the supported compute capability.

I don't own any of the cards. I am going to buy one when I decide which one is compatible and is best for my application.
    
CPUMotherboardGraphicsRAM
i7 950 ASUS Sabertooth GeForce 9500 GT 12GB Corsair Dominator 
Hard DriveOSMonitorPower
Cruical 64GB SSD / Seagate 7200 1TB Windows 7 Enterprise (2) Viewsonic 23.5" HDMI Corsair 850TX 
Case
HAF 932 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
i7 950 ASUS Sabertooth GeForce 9500 GT 12GB Corsair Dominator 
Hard DriveOSMonitorPower
Cruical 64GB SSD / Seagate 7200 1TB Windows 7 Enterprise (2) Viewsonic 23.5" HDMI Corsair 850TX 
Case
HAF 932 
  hide details  
Reply
post #4 of 7
Oh, in that case any card with a gt200 or gt200b core should be fine. I know my GTX 275 was 1.3.

EDIT: Unless of course you want something stronger with more up to date cuda compatibility, then fermi is an option.
Edited by malayalam - 3/3/11 at 4:02pm
Walrus
(13 items)
 
  
CPUMotherboardGraphicsRAM
Q6600 Asus P5Q-E Gigabyte GTX 470 SOC  Kingston 4 GB (Cheap and generic...) 
Hard DriveOSMonitorPower
Crucial M4 64 GB + Seagate 160GB 7200rpm + WD 1... Windows 7 x64 Pro / Xubuntu 11.10 x 64 ASUS VH226H 21.5" 2ms 1920x1080 Corsair TX750 
Case
Cooler Master 922 
  hide details  
Reply
Walrus
(13 items)
 
  
CPUMotherboardGraphicsRAM
Q6600 Asus P5Q-E Gigabyte GTX 470 SOC  Kingston 4 GB (Cheap and generic...) 
Hard DriveOSMonitorPower
Crucial M4 64 GB + Seagate 160GB 7200rpm + WD 1... Windows 7 x64 Pro / Xubuntu 11.10 x 64 ASUS VH226H 21.5" 2ms 1920x1080 Corsair TX750 
Case
Cooler Master 922 
  hide details  
Reply
post #5 of 7
Quote:
Originally Posted by Bagwej;12601016 
I don't own any of the cards. I am going to buy one when I decide which one is compatible and is best for my application.

Anything since the GT200 architecture would work.

If you want to go for used GTX260 or higher. If you want to go new you can get a good deal on either a GTX 460 or GTX 560 most of the time.
Edited by Lampen - 3/3/11 at 4:04pm
Singularity
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 6700K ASUS Z170-Deluxe GTX 770 SLI 6 GB Corsair Dominator GT @ 1600 MHz 7-7-7-20 
Hard DriveCoolingOSMonitor
Corsair Force LX Corsair Hydro Series H110i GTX Windows 10 Home Premium 3 x ASUS VX279Q 
KeyboardPowerCaseMouse
Corsair Gaming K95 RGB Corsair RM850i Corsair Carbide Series Air 540 Corsair Sabre RGB 
  hide details  
Reply
Singularity
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 6700K ASUS Z170-Deluxe GTX 770 SLI 6 GB Corsair Dominator GT @ 1600 MHz 7-7-7-20 
Hard DriveCoolingOSMonitor
Corsair Force LX Corsair Hydro Series H110i GTX Windows 10 Home Premium 3 x ASUS VX279Q 
KeyboardPowerCaseMouse
Corsair Gaming K95 RGB Corsair RM850i Corsair Carbide Series Air 540 Corsair Sabre RGB 
  hide details  
Reply
post #6 of 7
Thread Starter 
I saw the tables but wanted to verify their accuracy. I have found the answer from the man himself.
Quote:
Furthermore single-only cards are being phased out by NVIDIA. All new graphics cards are compute capability 1.3 or greater (the latest Fermi architecture is 2.0 and 2.1 and provides very good single and double performance).

About half way down the page
http://www.mathworks.com/matlabcentral/newsreader/view_thread/291689

Anyone around here use GPUs for what I am going to?
    
CPUMotherboardGraphicsRAM
i7 950 ASUS Sabertooth GeForce 9500 GT 12GB Corsair Dominator 
Hard DriveOSMonitorPower
Cruical 64GB SSD / Seagate 7200 1TB Windows 7 Enterprise (2) Viewsonic 23.5" HDMI Corsair 850TX 
Case
HAF 932 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
i7 950 ASUS Sabertooth GeForce 9500 GT 12GB Corsair Dominator 
Hard DriveOSMonitorPower
Cruical 64GB SSD / Seagate 7200 1TB Windows 7 Enterprise (2) Viewsonic 23.5" HDMI Corsair 850TX 
Case
HAF 932 
  hide details  
Reply
post #7 of 7
The GPU functionality in PCT is a joke. I started using it but was quickly disappointed. Luckily, I stumbled upon a tweet that pointed me to Jacket (http://www.accelereyes.com). Those guys have done a really good job - my code now runs 18X faster than it did with PCT. I can also using it with my CUDA 1.0 card in my laptop for when I'm on the move.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: NVIDIA
Overclock.net › Forums › Graphics Cards › NVIDIA › GPUs and Matlab