Do you remember when computers would “freeze” when you asked them to do too many things at once? Sure today’s computers are laggy at times, but they don’t freeze like computers of yesterday did. So that means computers have become faster, right? Well no, not really.
There are three terms you should know if you want to know how computers have became faster: clock speed, cores, and parallel processing.
Clock speed is the “velocity” in which your computer can think, and is measured in hertz (like gigahertz processors.) So, the clock speed has increased, right? Again nope, not really.
The “top” clock speed in today’s computer chips is in the 3-4 gigahertz range. The reason that this is the “top” is because of diminishing returns—as the gigahertz of the processor goes above 4, it takes ever more electricity to achieve higher clock speeds. Further, the challenge of heat and processor-damage become an issue.
But computers have become faster. So how have they become faster?
When people hear “computer chip” they think “one”. Indeed it is one chip, but it most likely has two or four “cores” on it. Cores are the “tiny computer chips” which make up the entire computer chip.
So why have computers become “faster”? Because they have more cores. Yesterday, when you would try to do two things at once, the first action would need to be pushed through the single core, and then the second thing would be pushed through the single core. Compare/contrast to today where each computer chip, with multiple cores, can compute multiple tasks independently of each other.
In a phrase, cores can be thought of as “the juggler being able to juggle more balls.”
Let’s say that you want to build a skyscraper. Let’s examine two methods you could use to build it.
- You could either build one floor, wait until it is built, and then build the next floor on top of it or
- You could preassemble all the floors on the ground, and then use a crane to stack those floors sequentially. Which method do you think is faster?
Parallel processing, the second method, is by far the-faster-method. The challenge with parallel processing, however, is that each floor must be “identical”. (Imagine a tall sky scraper. Tall sky scrapers usually stack the wider floors on the bottom and the narrower floors on the top. Because one floor is wider/narrower than others, the floors are not “identical.”)
Parallel processing is how “super computers” are built. But it also is how computers of the future will be built.
Your computer has became “faster” because it was able to juggle more things at a time because it had more cores to juggle with.