I am using ImageMagick to process large TIFF images, and I am running into memory issues as the processing happens in a limited resource environment (Function Compute on Alibaba, which is almost identical to AWS Lambda). Sometimes the process runs out of memory, and sometimes it runs out of temporary disk space.
I believe I am currently using a 64-bit version of ImageMagick's convert binary. If I were to custom compile ImageMagick binary under 32-bit, is there a change my image processing might use half the memory?
I believe the answer is no, but I just want to double check. My guess is that the convert binary loaded into memory will simply be half the size, and the actual image processing will use the same amount of memory as when using a 64-bit version of convert.