rogerdahl / cuda-collatz Goto Github PK

Hello Roger,

I found this project very interesting. After not doing C programming for over 20 years my old friend asked me to join him in doing a Collatz program as a hobby to while away the hours.

We are actually doing the max N value for any given N, rather than the delay. On my 4 core cpu I am up to 60, with no sieving, but using general number optimisations plus threaded so using 3 of the 4 cores ( it is my main PC so it needs to do something else as well on the fourth core, like write this :) ).

Being very rusty at programming, but having seen your excellent effort, I wondered whether you could advise on moving my code ( I use generic Visual Studio v17 C code ) to the graphics card; note I am an an absolute beginner at graphics card programming. The advantages are clear of course, graphics cards being massively multi-parallel. I have a 970 GTX in the machine not doing much nowadays apart from the odd game.... getting old :)

I am not sure if you are reading this page anymore, but I thought I would reach out on the off chance.

All the best, and well done on your work with this. I am always in awe of people who are on a higher rung of the programming ladder. Same for the maths bods !!!!