I looked in the TI-74 manual to see if it had anything to say about the SQR() function, and it only mentioned that it was the same as N^.5.
So I tried running the same program on the 71, 75 and 74 using N^.5 and interesting results. Both the 71 and the 75 took longer, but the difference was smaller, each came in at about 1 minute. Still faster than the 74 running SQR().
I expected that the 74 would take about the same amount of time, or less, thinking that the SQR() function just exponentiated and that was why it was slower. Guessing maybe they were trying to save code space or something and didn't have a separate algorithm for finding a square root.
But I was wrong. The 74 took almost six minutes! Three times as long! I thought something had gone wrong, but I kept waiting and it finally reported "Done!" with the right answer.
So now I'm going to test exponentiation versus just multiplying N by N and see which is faster. On the Apple II, it's faster just to have N multiply by itself, than raise it to the power of 2.
The things that entertain me...
Originally posted at Nice Marmot 15:39 Saturday, 2 September 2023