Overflow and Underflow Errors

[@KidsLearnHTML]

Big question, with today's powerful CPUs, is it possible for an overflow error happen in 2022? Any response would be greatly appreciated!

[@MrBrownCS]

Yes, I don’t see why this couldn’t still happen. But for these fundamental errors, there’ll be code/ circuitry within the CPU to flag it

[@RossMcgowanMaths]

Am I right in saying that we cannot get an underflow from floating point addition or subtraction. So for example the smallest 32 bit subnormal number is approx 1.4...e-45. So if I add or subtract any other normal or subnormal number I either get zero (1.4...e-45 - 1.4...e-45 = 0 ) or another subnormal or normal number. I can never get it to underflow ??? This seems reasonable to me but I cannot find any literature that mentions it in these simple terms.

[@wafikiri_]

You are right. Underflow from floating point addition or subtraction would use the tiniest of absolute values. You'd get underflow by multiplying too low an absolute value by too lower than one an absolute value, or by dividing too low an absolute value by too higher than one an absolute value. The result would be computed as zero but only a zero factor or dividend should yield a zero as product or quotient, thus, it's an underflow if that's not the case.

[@roasted_guava5706]

Do you have vids on CIE A levels? What do you think they will ask in the PAPER 3 of 9608 this friday?

[@cadetsparklez3300]

me jsut trying to figue out if oging to -217 etc is over or under: 8 minute video skull emoji