Here are some algorithms involved, and I don't understand them either. But some are easy to understand. For example, file A is placed on 10 clusters (I don't know how to pronounce "簇" exactly; I can't type it out), and file B is placed on 8 clusters. The positions of these two files on the disk are adjacent. We first open A and then open B. If the system reads by itself, it will read A first, and only when we are about to open B will it read B. But smartdrv will load both A and B into memory. In this way, the speed of opening A may not be much faster, but the speed of opening B will be much faster.
You can do an experiment:
Boot with a DOS boot disk, then type the following commands:
cd\
Press Enter
dir/s
Press Enter
At this time, the system will list all non-hidden files in this directory. Calculate the time required;
Then run smartdrv, type the above command:
cd\
Press Enter
dir/s
Press Enter
Calculate the time. Is it much faster than when smartdrv is not run?