Subtitles of the Movie

If someone were to ask you what eight divided by two is, I am sure you could quickly give them the answer four. But, how did you do that? Well way back in grade school using Flash cards or some other kind of memorization technique, you formed a mental picture in your mind that 8 divided by 2 is 4 - that's how the human mind works: mental pictures. But that's not how computers work. I a computer wanted to calculate the result of 8 divided by 2, the computer would start with the number 8, and starts subtracting the number 2. So that computer would say 8 minus 2 - that's 6. Minus 2 is 4. Minus 2 is 2. Now it took me three transactions to get there, plus the incidence itself of 2; so 3 plus 1 is 4 - that's how computers work. So a computer is very good at doing the same thing over and over and over, and that's why computers appear to be faster than people. In reality, people are much smarter than computers: people program computers. But computers are very good at doing the same thing over and over and over again very quickly. So since a computer is going to be doing the same thing over and over again, then why not take the resources that it uses. And once it acquires them one time, why not put those resources very close for the future. In other words, if the computer has to get some information off of a hard drive or some other source, that's fine the first time. But after that let's store it very close to the Processor so that it can use it again, over and over and over again in a much faster fashion. This is what Processor Cache does. We use high speed static RAM, special RAM that's more expensive and it's actually larger, but it's only used for cache. And it's used specifically to speed up processing working on the principle that computers do the same thing over and over and over. And speed and size are both important factors in regards to cache. As a matter of fact, cache must be balanced with the Processor. There are two different types of cache: these are L1 cache and L2 cache. L1 cache is contained within the Processor itself - so you can't get much closer than that. L2 cache is either in a component with the Processor, as in the Pentium chips that have a Processor and an L2 cache chip within the same single edge cartridge module, it's about the size of a whiteboard eraser. Or L2 cache can be located on the system board and connected by a very fast Bus. Now, we will talk about the Busses later on. But the speed and the size of the cache need to be well balanced with the Processor. In fact it's possible to increase the Processor speed and decrease the performance of a computer, if we don't increase the cache along with the Processor speed. This is because we can create a bottleneck. L2 cache is the helper. If the Processor all of the sudden needs a tremendous amount more help because it's faster, but the helper isn't up to the task, then we in fact have created a bottleneck within our own system. So it's very important to balance Processor speed with cache size and speed. And this balance can usually be obtained by just following the manufacturer's specifications for a particular system board or particular Server. Another way to increase processing speed and efficiency is by using more than one Processor. This is sometimes the best thing to do depending on the applications that we are using. When we use more than one Processor, and we control the Processor tasks, this is called Symmetric Multi-processing, and we will talk about Symmetric Multi-processing in our next section.