The phenomenon that leads to all of this is called the "skin effect".
Basically, DC current flows through a wire evenly, whereas AC current tends to want to flow along the outer edge (the "skin") of a wire.

If you concentrate current through a smaller cross-sectional area of wire (e.g. the skin) you end up with more effective resistance and greater losses. Thus, all other things being equal, you will lose more AC energy in a given wire than you will DC energy — the AC voltage drop will be larger.
Whilst the number of strands in a cable makes no difference for DC (only the total cross-sectional area, or gauge, matters), it does make a difference for AC. Cables with more strands have a higher ratio of "skin" to "core" and thus offer less resistance to AC currents.
That said, the skin effect is only really noticeable at high frequencies. At 50–60Hz in copper, the "thickness of the skin" is about 8–9mm. That's not cross-sectional area, that's depth (δ). So for skin effect to have any appreciable impact at typical household frequencies, the cable diameter would need to be 16–18mm thick. That's a massive cable that could carry several hundred Amps — not at all what you are probably using to hook up the average PV array.
Long story short: 'DC/AC Cable' (solid/stranded) doesn't matter at all for DC links, and doesn't really matter for low frequency (50–60Hz) AC links running through cables thinner than your fingers.
Multi-strand cable is, however, a lot easier to bend — so if you want/need a bit of flexibility then increase the number of strands.
If you've got microinverters on your panels (e.g. Enphase), and AC coming from your array, and want the "Done Right" achievement badge, by all means — go nuts on the number of strands. You're probably going to see less than a 0.1% improvement, but it will make you feel better. At such frequencies, the advantage of 'AC Cables' is more 'theoretical/nominal' than 'practical/actual'.