I really love the program (tried 4.6 and 4.7) but it runs very unstable on my system (i7 2700k, 16 GB ram, Geforce 550 ti 1GB). I use about a 1000 tiff files of 800 kb each.
Am I asking too much of my system, and therefore get the “not responding” message very often?
Do you maybe have a solution? At the moment it is very difficult to work with.
Crashes and hangs are extremely rare in recent versions of Slicer and they are almost always due to running out of memory.
Probably your tiff files are compressed, so your raw data set size may be 2-8GB. Typically, you need 10x more memory space than your data set (as you need memory for intermediate processing results and temporary variables), so 16GB memory space will not be enough.
You have to increase your virtual memory size to about 60-80GB (backed up with as much physical memory, as possible, to get better speed) and/or reduce the size of your dataset (crop, resample, convert to grayscale, etc).
I am monitoring RAM usage and it typically does not go above 10 GB. When it says “not responding” and I monitor cpu, crashes with 10-20% CPU usage resolve themselves (its just slow), crashes with 90-100% kill the program.
Nevertheless I will try the above suggestions and see if it works. In the near future I may also have the possibility to run it on a higher-end system (32 GB ram), maybe that resolve these issues too.
Would you say that the GPU plays a role as well? I had it enabled and set to use 1GB of the 1GB on the card. I also disabled it (i though that maybe there was a problem with drivers e.g.) but the problems persisted.
Note that you don’t need more physical memory, just make sure you allocate enough virtual memory/swap space on your disk. Backing it up with as much physical memory is possible, because disk access is slower than RAM.