Why does
Two CPU’s != Twice the power
In: Technology
Rule 2: motivations of businesses and groups
Throwing more cores or CPUs at the problem only works if your problem is that the computer can’t perform enough calculations at the same time.
For most tasks performed by personal computers, there are logical reasons why you can’t perform multiple tasks at the same time (in parallel) and must perform them one after the other in sequence (in serial). And in those cases you can only perform the task quicker if you can do the individual calculations faster. If your job is to add 2+5, and multiply the result by 10, you can’t do the multiplication by 10 without knowing the result of 2+5.
If you and your housemate are driving from your house to a restaurant, you won’t get there in half the time if you take separate cars.
Having multiple cores on a single CPU is fine for consumers who only have a handful of programs open at a time, most of which need to do serial processing. In some certain cases, like back-end servers for web sites and business software, virtual machines and such, they can make better use of more threads and CPUs. Which is why server hardware often supports two or more CPUs. But even then it only works to a certain point, because those CPUs also have to share the other system resources like memory, storage, GPU, networking etc which can only handle so much processing/data, and all those components talk with each other via the motherboard’s various [buses](https://www.webopedia.com/DidYouKnow/Hardware_Software/computer_bus.asp), which also have a limit as to how much data can flow.
There’s a concept called parallelization, which is how well a task can be split up into completely independent tasks. Let’s look at two sets of math calculations run with two CPUs
x=y+3
z=2*x
We assign the first equation to CPU 1 and it gets to work. However, we can’t assign the second one to CPU 2. This is because the second equation needs the answer from the first equation. So CPU 2 doesn’t do anything.
Let’s look at a second set.
x=y+3
z=2*q
We assign the first equation to CPU 1 just like before. Now equation 2 doesn’t require equation 1 to be solved first. Now we can assign CPU 2 to run equation 2 at the same time.
We can extend this concept to more general tasks. Some tasks benefit from more CPUs and some don’t. Play music and watch a video at the same time? parallel, 1 CPU plays music and the other plays the video. Run a physics simulation of a bouncing ball? Not parallel (simplifying). You can’t calculate what happens at second 10 without finding what happens at second 1.
So short answer is they have and do but it’s basically secluded to cpus such as xeon so server based and server based motherboards.
Technically you can do anything with this hardware such as playing games but it’s primarily used in servers.
They can and do. But for the average consumer, the cost quickly outweighs what a consumer needs and is willing to pay for.
My computer is dual socket, which means it has two cpu slots on the motherboard. Currently I only have a single $2000 cpu in there, but when I need more I’ll buy a second one.
Currently I edit raw 4k footage and do 3d rendering, and one is plenty.
Why don’t you read a book twice as fast because you have two eyes? One eye per page.
Latest Answers