A beautiful day to all primers,
i have develloped a prime sieve base on Ulam for a Cuda Graphic Card.
constant ca 8.317.200 Primes / sec
on a Cuda graphic card Geforce 650 with 192 Cores
there is a description of the algorithm and an implementation
Perhaps there are some speedups possible.
The algorithm is quite nice and i hope it is possible to understand the idea.
I am interested in results on graphiccards with more Cores
Nice greetings from the primes