I’m not sure if it’s that much faster in double precision without a high end Tesla A100 or H100 that actually has the TFLOPS of FP64. Yes in single precision it is lightning fast. I would LOVE to see them get the multiphase solver ported next!
@OzenEngineeringSimulation Жыл бұрын
Just tested it on our machine. Double precision is also blazingly fast on our RTX4050. Maybe 60% slower than single precision and still way faster than CPU.
@ayomasukaja9 ай бұрын
@@OzenEngineeringSimulation please review 4050 vs i9 vs ryzen 9 or what you have or old tesla k80 that has double precision
@tea-hash-sea4 ай бұрын
@@OzenEngineeringSimulation So, do you confirm that Fluent can run with Geforce cards ? (both native GPU solver and acceleration of CPU solver ?)
@JA-sj9xk Жыл бұрын
Hi, I got myself a quadro rtx a2000 and the utilisation is 0% on simple method, I have a r9 3950x 16 cores with 64gb ram, any idea why it doesn't work?
@tomasmartins3541 Жыл бұрын
Your probably not making sure to select the GPU to run calculations. There are videos on YT to show you how to make sure they are working