Error is introduced by use of Paint and MO itself, opening and closing them at each stage may affect the results a bit. This was not a scientific experiment by any means but it shows that the tool does work. I took data and pictures of before and after for these steps:ģ) Optimized after Steam/Launcher 4,421,204Ĥ) Vanilla Arma 3 loaded on a LAN server on Tanoa 1,862,916Ĥ) Optimized after Arma 3 loaded 2,960,760 (!)ĥ) Arma 3 unloaded (Steam still in memory) 3,789,860

It took about 30 min to do a test using MO in various stages of Arma3 play. Maybe there are tools out there that can. I wish MO showed graphically how fragged the RAM actually is and what it looks like after instead of that dumb circly thingo. Snake oil? Do I have the fundamental understanding that the RAM can get fragmented? Does this indeed cause malfunction? If the RAM is kept tidy, is there less of a chance of this happening? I say yes. Or it could oversteps its bounds if poorly coded and that causes all sorts of malfunctions, (0xc0000005 errors for example). This allocated memory can become chopped up into small segments over time, sometimes too small to use so when a program needs a block of memory for some more data it cannot find a chunk big enough to continue.

Each program has memory assigned (allocated) to it that it, and only it, can use. What the program does is de-fragment your RAM.

BTW, I run it right after Arma is loaded and the launcher closes and has released it's memory back to the system. I use it whenever I please, even when Arma is running, between scenarios, etc. I don't think Memory Optimizer helped but I still use it because I understand the premise of fragmented RAM. Perhaps BI fixed the bug that was killing my instances. Yeah, I suggested it only because I have tried it to overcome the 0xc0000005 errors (memory access violation) I was getting constantly.

How does wise memory optimizer work