What are quad/dual cores used for?

638 views

I know its supposed to be able to do multiple tasks at once, but what are these task? For example, is it like playing video games while listening music or something. Im sure its probably a lot more advanced than that, so please give examples.

In: Technology

5 Answers

Anonymous 0 Comments

yes, playing video games while listening to music is multitasking.

on a single processor it would just split up the time among each task, since the cpu doesn’t need to be utilized all the time for each task it can split up it’s time among various tasks. then on a multicore, each task can get it’s own cpu which allows more tasks to run at any one time.

Anonymous 0 Comments

Most modern programs are built to make use of multiple cores. Instead of designing each process to consist of multiple steps to be taken one after the other, it is divided up into tasks that can be all done at the same time.

The reason for dual or quad cores (or more – the leading AMD cpu gives you 32!) is that our ability to make computer chips run faster is limited by physics, but our ability to pack more stuff onto a chip keeps on getting better. So we pack multiple CPUs on one chip, and build programs that take advantage of those multiple CPUs.

Anonymous 0 Comments

The computer uses all the cores to execute your running applications. The more cores the more that can be run in parallel.

Picture your computer like a grocery store. Your applications are tasks like shoppers queueing up at the checkout to execute. Some task are simple (not much in the cart), others are more complex (more in the cart). If there is only one cashier then shoppers (tasks) get backed up waiting. If there are multiple cashiers then they can unload (schedule) more shoppers and there is no backup and no applications wait very long for their task to execute.

Anonymous 0 Comments

It’s obviously more complex than this, but yes, it’s your device multitasking.

Computers run a *lot* of tasks at the same time. Even if you have no programs open, there’s a ton of stuff going on in the background.

Your processor has to schedule time for all of this stuff to happen, and it’s actually extremely good at it. The faster the clock speed (how many things it can do in a second), the faster it can get stuff done. There is a limit to this, though, both how fast a processor can go and how much it can handle.

But never fear! We invented dual-core processors! The processor got its twin a job, and can take some of the workload off its hands. They schedule what’s best for them to do by themselves, then do it. Suddenly the workflow is more efficient and gets done faster, freeing them up to do more tasks.

But what if the workload is still too much? Never fear, for they’re quadruplets! Or octuplets, or… sexdecuplets… because we now have 16-core processors (and more!). Imagine how much can get done with that many processing cores!

Some programs can ask for multiple processing cores to help out with a task, since the workload is so extreme. Games take advantage of this, like having one core deal with artificial intelligence and another with the physics engine.

And sometimes, these -uplets have an assistant who can help out each processing core and do some of the work for them. This is called hyperthreading, which is why you’ll sometimes see processors advertised as “4 cores, 8 threads”.

Anonymous 0 Comments

Your CPU currently checking if you moved your mouse, it’s checking if mouse has been clicked, it’s checking what’s under the mouse and sending a signal to a program that’s responsible for the window that the mouse is on about mouse being there. It’s also receiving the signal as the program whose window is there, deciding if it needs to change something due to mouse position. Program also might be doing something else, like making requests to web servers, and CPU handles that too…

With single-core CPU, all those myriard tasks are done sequentially. Processor however swaps between tasks very fast, like, once every millisecond at least, or more, so it seems like everything is done simultaneously, but actually all those tiny things happen by processor switching tasks.

Even before multicore processors, it was found out that it was good idea to logically separate these tasks into first programs(so each program got its own slot), and later this was refined even further to be threads. Each program basically got onto a queue, and when it was their turn, they had like a millisecond to do things on CPU. After the time was up, or program voluntarily gave up CPU, next program would come in, and the program would get back into the queue. Operating systems would manage this. CPU only did one thing at once, but because of constant swapping, it seemed like your multiple programs were running simultaneously.

This was further developed into threads, so single program could kinda have “mini-programs” in it, like, if you calculate trillion digits of pi and show a cute dancing triangle that made a sound, while waiting for the computation to finish, you could have one mini-program calculate the digits of pi with no worries, and second miniprogram to check every couple of milliseconds if you have clicked the triangle so the program can make the sound. It’s a bit of a mess but the basic idea is that even just one program can be split into many logical tasks that can be handed to the CPU separately. If that doesn’t make sense to you, then you’re not missing much if you just ignore threads and think of it all being programs in a queue.

And now, with multiple cores, you can split the queue. So instead of just letting one program use processor at a time, you can now have multiple programs do their stuff simultaneously, so theoretically you could have computer get 4x faster by having 4 core processor.

But also, if all your computer runs is a single program(single-threaded), then having 4 cores is not any better than 1 core. That program gets to the CPU and the individual core presumably isn’t any faster at executing it than the mono-core processor. But modern computers tend to run so many programs the same time that you’re guaranteed at least some speed boost from multiple cores. How much? It’s hard to tell exactly.