Do you know if the new Nvidia’s system DGX Spark will be an efficient and usable tool to accelerate molecular simulation computing with gromacs or other software? Thanks for your answers and/or comments.
If this simply uses a scaled down Blackwell chip, the FP32, and thus GROMACS, performance is going to be far less than what you can get with a gaming system with one or two consumer GPUs for the same price or less.