Hi Nadav,
I don’t believe any benchmark tests have been conducted in-house. For a general comparison between the two processors, there are a number of benchmark websites, such as “userbenchmark.com”.
Regarding RAM utilization: 32Gb should be enough for most types of simulations. RAM usage typically ramps up when the system is optimizing a very large number of non-sequential rays through a complex system and/or has imported CAD components. For these cases, 32GB should be OK, but maximizing memory on the machine would be recommended.
Here is a community post, where the user ran into insufficient memory for reference:
Post this to the Reddit optics group. There’s someone there who will be able to answer this.
To the Zemax team: It would be great to get some detailed information on this. The Intel CPU mentioned above has 8 “performance” cores an 16 “efficient”. The AMD just says 16 cores.
Which one is faster? Also, does the answer depend on sequential vs non sequential?
I don’t know about Intel’s marketing these days, but I distinguish between ‘real’ cores, which are actually part of the silicon, and hyperthreaded cores, which use a scavenging approach to make unused CPU cycles appear as a separate core. They’re software cores, rather than real. Real is best.
Based on that page, my guess is that Performance Cores are what a big number-cruncher like OS would use, while the Efficiency Cores would run smaller tasks like system-level processes. But it would be fun to play with one of these big boys!