The task is to write C++ prog for some of the TPCH queries... in a sequential way as per the query dissecting the primitives. Then write them again into OpenCL for parallel execution using Kernels.
The whole project is to evaluate the performance of TPCH queries using OpenCL on GPU
(*****Either u can directly write the code in OpenCL or in C++ and then into OpenCL it's up to your flexibility*****)
I can send you some sample files of OpenCL for consideration
Also i have query plans for some of the TPCH queries