I am assembling a computer to run my Abaqus explicit simulations as fast as possible. As far as I know, the CPU and RAM are the most critical components. Based on my experience, parallel processing with more than 24 CPUs doesn't reduce the runtime of my simulations, so I am prioritizing the clock speed of the CPU. Not sure whether the clock speed or the cache size is more important. For the clock speed, I have selected the Intel Core i9-13900KS. My simulations typically involve fewer than 100,000 elements, so 128 GB of RAM should suffice. For this, I have chosen 4x32GB Intel XMP, CL 40, 5200MHz modules. Not sure about the impact of the graphics card since GPU acceleration is not utilized in explicit simulations. Would it make a significant difference in implicit simulations? Also not sure about the extent to which the motherboard can affect performance. Dell offers a workstation (Dell Precision 3660 i9-13900KS/128GB/2TB/A2000) that, in my opinion, is one of the best options available for my simulations. Probably anything more expensive wouldn't significantly affect the runtime of my simulations.

About my models: I employ very complex user material and damage models, with numerous parts in contact at a system level over extended durations, around 500 seconds. I run all simulations in double precision with full nodal output precision.

Budget: up to \$10,000.

Abaqus ​​​​​​​