GTX690 + CUDA
Hi all,
how many CUDA device the system will see using a GTX690?
If the system sees 2 CUDA devices would it mean that the two GPU do share the
on board Memory (no need to memcpy)?

Regards
Hi all,

how many CUDA device the system will see using a GTX690?

If the system sees 2 CUDA devices would it mean that the two GPU do share the

on board Memory (no need to memcpy)?



Regards

#1
Posted 04/30/2012 11:46 PM   
CUDA will see two independent devices, each one with its own 2GB of memory.
You will still need to memcpy, but using p2p you should be able to use the internal gen3 switch and achieve pretty high transfer rates.
CUDA will see two independent devices, each one with its own 2GB of memory.

You will still need to memcpy, but using p2p you should be able to use the internal gen3 switch and achieve pretty high transfer rates.

#2
Posted 05/01/2012 02:18 AM   
Can the two CUDA devices still communicate at gen3 speeds if plugged into a gen2 motherboard?
Can the two CUDA devices still communicate at gen3 speeds if plugged into a gen2 motherboard?

#3
Posted 05/01/2012 02:10 PM   
Yes, if you are using P2P.
Yes, if you are using P2P.

#4
Posted 05/01/2012 02:28 PM   
Excellent!
Excellent!

#5
Posted 05/01/2012 03:16 PM   
[quote name='mfatica' date='01 May 2012 - 07:28 AM' timestamp='1335882537' post='1403027']
Yes, if you are using P2P.
[/quote]
What's the API interface for that?
[quote name='mfatica' date='01 May 2012 - 07:28 AM' timestamp='1335882537' post='1403027']

Yes, if you are using P2P.



What's the API interface for that?

#6
Posted 05/01/2012 09:05 PM   
[quote name='RezaRob3' date='01 May 2012 - 01:05 PM' timestamp='1335906314' post='1403165']
What's the API interface for that?
[/quote]
cudaEnablePeerAccess/etc.
[quote name='RezaRob3' date='01 May 2012 - 01:05 PM' timestamp='1335906314' post='1403165']

What's the API interface for that?



cudaEnablePeerAccess/etc.

#7
Posted 05/01/2012 09:32 PM   
[quote name='tmurray' date='01 May 2012 - 02:32 PM' timestamp='1335907939' post='1403177']
cudaEnablePeerAccess/etc.
[/quote]
The Best Practices Guide says, "UVA is also a necessary precondition for enabling peer-to-peer (P2P) transfer of data...," but I can't find any information on UVA in the Programming Guide or the Toolkit Reference. It doesn't even say that UVA is related to P2P! What am I missing?
[quote name='tmurray' date='01 May 2012 - 02:32 PM' timestamp='1335907939' post='1403177']

cudaEnablePeerAccess/etc.



The Best Practices Guide says, "UVA is also a necessary precondition for enabling peer-to-peer (P2P) transfer of data...," but I can't find any information on UVA in the Programming Guide or the Toolkit Reference. It doesn't even say that UVA is related to P2P! What am I missing?

#8
Posted 05/01/2012 10:05 PM   
you need to be running a 64-bit process in order to enable P2P, basically. also not WDDM.
you need to be running a 64-bit process in order to enable P2P, basically. also not WDDM.

#9
Posted 05/01/2012 11:22 PM   
Scroll To Top