SimFlow Hardware Recommendation
Recommended resources in terms of CPU, GPU, RAM, and HDD (Hard Disk Drive) for running an analysis in Autodesk CFD.
RAM memory is used to store simulation data and software instructions that the CPU needs to perform the calculations. The amount of RAM you need for CFD calculations will depend on the size and complexity of your simulations. For most simulations, where the number of nodes is below 10M, 32GB RAM will be well enough. However, if you will run multiple analysis on the one PC, it may be worth to increase the memory up to 64 GB. For the larger models above 10M nodes, we will recommend about 2-3 GB of RAM per 1M nodes.
Whether the simulation require more memory than it is available on the machine, the data will be stored on the hard drive. It will significantly slow down the calculation and increase its time. In this case we recommend to increase the RAM memory. The RAM speed has a minor impact on the calculation speed. If the RAM memory size is not limited for evaluated case, the RAM speed can have an impact on the time of calculation. Faster RAM can speed up calculations by a few percent.
When it comes to CFD calculations, the performance of the CPU is critical to achieving fast and accurate results. CFD computations, like most CAD computations, are memory bound, thus the best CPU for this task would be some balance between the highest number of memory channels and the number of cores.
Here are some general CPU recommendations for CFD analysis:
- Choose a processor with a high clock speed – this is because CFD calculations are often single-threaded, meaning they can only utilize one CPU core at a time. So, a CPU with a high clock speed will be able to perform each calculation faster.
- Consider getting a processor with multiple cores – if you plan on running multiple simulations at the same time or doing simulations that require multiple cores, then having a processor with multiple cores can significantly improve the performance of your CFD calculations.
- Look for a processor with a large amount of cache – cache is a type of memory that a processor uses to store frequently accessed data. A CPU with a larger cache can help speed up CFD calculations by reducing the amount of time it takes to retrieve data.
- Consider getting a processor that supports hyper-threading – hyper-threading allows a single CPU core to work on two tasks simultaneously, which can improve the performance of CFD simulations that utilize multiple threads.
In CFD computations there is a lot of data transferred from memory to the CPU, and usually, the CPU is fast enough to compute but a lot of time waits for new data. Therefore, a larger cache and the number of memory channels (and so memory bandwidth) play a significant role and may become limiting factors.
Some specific CPUs that are often recommended for CFD calculations include the Intel Core i9 and Xeon processors and the AMD Ryzen and EPYC processors. However, the specific CPU recommendation will depend on the type of CFD simulation you are performing and your budget.
A graphics card with 2 GB should be adequate for most types of analysis. However, if you are working with larger analyses and experience slow interaction with your models,
upgrading to a 4 GB card could be beneficial. It’s not necessary to invest in an expensive Quadro card, which is designed for CAD but may not be worth the extra cost. Instead, a good gaming GPU card should suffice. In general, any nVidia card with 4GB or more memory (depending on your budget) should be sufficient for post-processing.
Physical Memory (Drive) Recommendations
Use a drive with sufficient storage capacity – CFD simulations can generate large amounts of data, so it’s important to choose a drive with sufficient storage capacity to store all of your simulation data.
Consider using an Solid State Drive – SSDs are significantly faster than traditional hard disk drives (HDDs) when it comes to reading and writing data. This can help reduce the time it takes to load and save simulation data, improving the overall performance of the simulation.
Choose a drive with a high read/write speed – CFD simulations involve a lot of data, so having a drive with a high read/write speed can help reduce the time it takes to access and save this data. Look for a drive with a read/write speed of at least 500 MB/s.
Consider using a RAID configuration – RAID (Redundant Array of Independent Disks) is a way to combine multiple hard drives into a single logical unit for improved performance, redundancy, or both. A RAID configuration can help improve the overall performance and reliability of your storage system.
In summary, a fast SSD with a high read/write speed and sufficient storage capacity is recommended for CFD calculations. If possible, consider using multiple drives or a RAID configuration to improve performance and reliability.
If the solver has to resort to file swapping (if the RAM limitation is reached) then Disc speed makes a larger difference.