In investigating the “excessive disk write” issue that has been in the news as late, I am seeing the very same thing happening on my M1 MacBook Air. It seems that with every image that is exported, 2 GB of RAM is used and just continues to grow with each export until RAM is used and disk swapping begins. It may be a possible issue with Rosetta and/or PL4 and/or Big Sur. Yesterday I only processed 11 D7500 Raw images and exported to 16 bit TIFF with Deep Prime. On the M!, PL4 consumed 9.5 GB of RAM (2/3 virtual). Because of the speed off RAM and SSD and tight coupling, no slow downs were noticed. So, I wanted to see if it was PL4 endemic so I ran the same 11 images with PL4 on my iMac with Mojave (Deep Prime was excruciatingly slow) and after the first image, PL4 only used 2.3 GB of RAM and stayed there. Now perhaps it did request more RAM, but if it did, it properly was returned to the system.
What is also interesting, after processing the first image on my M1, I saw an identical size increase, except on the M1 it just continued to grow. So, perhaps
the Rosetta translator is not relinquishing the RAM even though PL4 requests it? I would think there should be some basic error checking in PL4 to insure it was returned to system (Dynamic Memory Allocation 101) and fault if it didn’t?