Hi everybody,
I am wondering if the refinement will be significantly faster by caching data on a SSD that has very fast random read/write rate, for example, a 1-2 TB M2 SSD card be used for a refinement, it will fasten the speed of refinement?
Another question, if I use 1TB RAM for our workstation, do you think it’s significant to increase the rate of data processing and refinement ?
any answer is welcome, and appreciate your attention.
Thanks,
Hi @Dali,
SSD caching does indeed make a significant difference to refinement speed and an M2/NVMe SSD will be best. Having more RAM does not directly make refinement faster but does allow for larger box sizes and particles. However, on most Linux systems, the RAM is also transparently used as a file cache, so with 1TB of RAM, the system may be able to keep recent particles in RAM which will be even faster than SSD. Overall i would say the cost of 1TB of RAM vs. 2-4TB of SSD will mean that the SSD is the way to go.
Hi @apunjani,
thanks for your reply. Sounds great. It makes sense! we’re planning buy a 2TB M2 SSD for our workstation. Thank you!