Hi, using an S9150 for crunching FP64 algorithms. Was previously using an 280x consumer grade card.
The performance of the S9150 is a little better than the 280x, bet no where near the expected 2.5x performance improvement.
Making this assumption based on the 280x having approx. 1Tflop of double precision vs. 2.5 of the S9150.
I'm thinking I must be doing something wrong. How can I get the desired performance improvement out of the S9150?
Any help would be much appreciated.