On Saturday, 5 August 2017 at 01:15:19 UTC, Michael wrote:
On Saturday, 29 July 2017 at 06:54:47 UTC, Prasun Anand wrote:
I wrote a Linear Mixed Model tool for Genome Wide Association
It is built on LDC
and is faster than its Python alternative. Also, its the only
GWAS tool with a GPU
I am interested in porting ` faster_lmm_d` to Ruby. Though, it
is still a work in progress, I have written a blog about my
I would love to hear your feedback.
[Previously posted on LDC thread:
I don't know much about your work but it sounds like an
interesting application for D. How easy was it interfacing with
It was easy interfacing CUDA. I used `dstep` to create the
for CUDA and CUBLAS headers. I got interesting results but would
on improving the performance.
Was it just cuda you targeted or would it also work with OpenCL?
Yes, OpenCL is supported using ArrayFire. I have targeted CUDA
for better performance.
What made you choose D over going straight to C to work
directly with cuda?