I like the top comment but I thought maybe I’d try another, quite relatable take:
Think of something like a county’s election. The most recent US one is still fresh for most people. You had around 155 million votes in that election. Those all need to be counted.
If you have one guy name Steve who has to count them all one at a time, it’s going to take a looooong time. So you hire 2 people. Or 3 people. Or 50 people. Or a million people.
Then you spread them around so that the votes are counted closer to their origin. Then you upgrade to machine counting, etc.
In the end you have one overall “tally” with *lots* of different pieces involved in the counting and done *much* faster than if it was just that one guy counting.
A super computer is similar in that it takes one *very* large and *very* complex task and completes it much faster than a single, simpler computer could. Your home PC is Steve. The super computer is the entire system of people counting.
In the end Steve could maybe count a few thousand votes on election night. But the entire system working together can give a pretty damn good picture of who won an election *on election night*. Much faster.
Latest Answers