The customer asked us to introduce a calculation server whose main purpose is to run the thermal fluid analysis solver STAR-CCM +, MATLAB, etc. with a budget of about 100 million.
At the time of your inquiry, the customer seemed uncertain whether to prioritize CPU or clock or parallel. Most of the CFD analysis software is specialized for parallel, so if there is little difference in the number of CPU cores to select, I think that there may be a choice to emphasize the clock, but if the difference in the number of cores becomes large Think that it is desirable to attach importance to the number of cores. Therefore, this time, we proposed a specification with as many cores as possible within the budget.
In addition, I was hoping for a video card that could process the analysis results of a little less than 1000 million analysis grids to some extent, so I chose NVIDIA Quadro M2000.
【Main Specifications】
CPU | Xeon E5-2680v4 (2.40GHz 14 cores) x2 (28 cores in total) |
memory | 64GB ECC REG |
storage | HDD 1TB |
network | Gigabit LAN x2 |
video | Quadro M2000 |
Housing + power supply | Tower housing + 850W power supply |
OS | 7 CentOS |
■ Click here for details and inquiries about this PC case Example of a machine with a budget of 100 million for CFD analysis |