A05 Solutions

Reference answers with reasoning. Try the ../exercises/ yourself first - the learning is in the struggle, not the answer key.

This directory has its own Makefile:

make all GPU_ARCH=gfx950      # build every solution for your GPU
make run                      # build + run all; every one must PASS
make clean

#	File	Notes
1	01_exclusive_to_inclusive_solution.cpp	incl[i] = excl[i] + in[i] (no boundary case) vs excl[i] = i ? incl[i-1] : identity (two traps: the identity at 0, and the in-place shift is a race). Grand total = excl[n-1] + in[n-1].
2	02_segmented_scan_solution.cpp	Carry the head flag with the value: if (have && sf[tid] == 0) { sv[tid] += add_v; sf[tid] \|= add_f; }. Includes why the pair (value, flag) forms an associative operator.
3	03_scan_analysis_solution.md	Full work/depth math, the (log₂n − 1)/2 extra-work law, barrier and utilization counts, LDS footprints, arithmetic intensity, and the three regimes where Hillis–Steele wins.