For HFT: Nvidia CUDA Visual C++ financial examples show power of GPU versus horrendous cluster hardware management
With the research below it seems that CUDA may be the easiest way to go working off the supplied CUDA examples for financial needs. There will be no need of clustering, servers, etc which will make life so much easier. The outstanding item is to see C++ CUDA examples with trading signals which I am now seeking.
1. Download CUDA toolkit https://developer.nvidia.com/cuda-downloads. I am using Windows 7
2. Import the samples into Visual Studio 2010 and look at the C++ Financial samples
Note on how CUDA is much faster than Open GL espeecially on GeForceGTX 480. Below is used for technical indicators.
The listed GeForce board is quite affordable as well for $340 for 480 cores. Nice!
Note: I have dropped the idea of Fastflow with Visual C++> The author did not bother showing how to use Cmake to build the library so I dropped the idea. I really don’t care about the benchmark anyhow as I just want to see it work as fast as possible.
See how I proceed with this to build my HFT platform by joining my newsletter
I now post my TRADING ALERTS
into my personal FACEBOOK ACCOUNT
. Don't worry as I don't post stupid cat videos or what I eat!