I am looking for an expert CUDA programmer to complete the following tasks in Windows. Please only apply if you very familiar. Please be prepared to do a coding challenge live.
Here are the tasks needed:
1) Sort a 1-dimensional array of double / float
2) Sort an 2 dimensional with n-colum array of double / float by a single dimension (rowwise or columnwise)
3) Pearson correlation coefficient of two columns in an 2 dimensional N-column array
4) Spearman correlation coefficient of two columns in an 2 dimensional N-column array (this basically consists of sorting the arrays and calculating the Pearson coefficient on the indices)
5) Kendall correlation coefficients of two columns of an 2 dimensional N-column matrix
6) Covariance of two columns of a 2 dimensional N-column array
7) Covariance matrix for an 2 dimensional N-column array
Final deliverable should be performant, memory efficient, and tunable to maximize utilization of a specific processor.
coding should be neat and clean, with a reasonable amount of well-formatted comments. Feel free to use additional libraries like cuBLAS to perform the tasks.
10 freelancers are bidding on average $605 for this job
I have used CUDA efficiently for some years and in various kind of algorithms. I am in position to undertake this series of tasks, we could discuss more about details and deadline Thanassis