A couple of days ago I posted a study by Bron that I found quite cute.
1.Bc8+ Kg3 2.Be6 A fork, forcing one of the knights to move, but they can both give check which allows a defence.
2..Nb4+ [2..Ne7 will fail to 3.Kd7 when there are no more useful checks, so one of the knights must fall] 3.Kc5 Na6+
Now that the a6 knight is protected, the other knight can be moved. So white gains time by attacking black's bishop, which must then move to a poor square. 4.Kd4 Bh7
At least both knights are now safe! However, the poor placement of black's bishop allows white to trade it now. 5.Bc4! gaining time by attacking black's unguarded knight on a6 5..Nb4 and the knight responds by protecting the d3 square which is where white's bishop wants to go.
And finally, white succeeds with the plan, by playing 6.Kc3! attacking the knight, but simultaneously eyeing the d3 square. 6..Nc6 The only safe square.
7.Bd3 now forces the trade of bishops, as black's bishop has nowhere to run, leaving the game with 2 knights vs a lone king, a draw.
While this might not seem that difficult, there is something of a beauty to the way it was composed, with white's bishop and king both able to gain time every move to improve their positions while black's bishop proved useful to start with in aiding the knight on the queen side, but ended up being the weak link in black's position.