We have run Benchmark tests on new LinuxVixion Workstation model LVX SILENT 2*RTX-3080 and compared with previous LVX-SILENT 2*2080Ti Workstations with Core i9 10940X.
We are installing the new Nvidia Ampere family of cards: Geforce RTX-3080 (10GB RAM, 8704 CUDA Cores, 1440 MHz Clock speed) and Geforce RTX-3090 (24GB RAM, 10496 CUDA Cores, 1395 MHz Clock speed).
Our GROMACS 2020.4 tests have obtained an increase in computing speed over the previous 2080 Ti, ranging from 12% in AGH DODEC PME to 37,5% in the case of ADH CUBIC PME. This Geforce RTX benchmark is without a doubt an improvement in GROMACS calculations.
As you can see in the following image, RNASE calculations also become faster with the new Nvidida Geforce RTX 3080.