Cluster computing - Tesla C1060
Gaempi ... First, no. Tesla card advantage with CUDA come only if you can call the parallel calculation architecture from these chips. And I think it;s not possible with EL only. With Matlab and the language and other languages it's very simple to call it. Otherwise you have to write something which exports your data or something like that, calculates it in CUDA and translate it back for using in MC.
If you want speed up MC to calculate something with huge access of tick- data and so on, you can install it on a RamDisk.
Mike
If you want speed up MC to calculate something with huge access of tick- data and so on, you can install it on a RamDisk.
Mike
- Dave Masalov
- Posts: 1712
- Joined: 16 Apr 2010
- Has thanked: 51 times
- Been thanked: 489 times