Buying PC

[Som1Lse]

Are you sure the clock rates don't relate to the numbers of instructions? I haven't got my information from anywhere directly; I just noticed a property of ASM instructions in multiple documentations and articles called "cycles". For instance, take a looks at this. The "MOV address1, address2" ASM instruction requires 2 cycles to complete. Doesn't this mean that a processor with 2.0 GHz could complete 1*10^9 "MOV address1, address2" instructions per a second? If so, then isn't this direct evidence that the clock speed, or its' frequency to be precise, what mostly makes a processor faster at processing processes?

On 8051 architecture yes. However there are a bunch of different architectures out there and all of them differ.

Definitely one can do as much threads as he likes, but they would only cause the increase in performance if each is run independently in one core. As far as I know, if a program uses more threads than the cores available, then the CPU would just run the different threads as if they were separate processes in the same core. This shouldn't cause any increase in performance, but further worsens it with the extra semaphores and synchronization stuff.

Yes, although there are times where using multiple threads makes programming easier than doing everything on a single thread (generally when little synchronization is required).
Other than that you are correct as far as I know.
I'll point out that a game will still not use one thread for logic, one for sound and one for networking. It really depends a lot on the game.

Sounds interesting. Just to make sure I got the idea of the cache memory right; I assume this is where the variables and stuff you don't store in a heap go? (I remember when I first started C++, my program threw a Buffer Overflow because of a non-new allocated array that slightly began to exceed 4MBs in size).

Not at all.
Cache is memory that is on the CPU itself. The variables you declare inside a function are not (necessarily) stored in cache, but are stored on The Stack.
When you are "new"ing up something then it gets stored on The Heap, but if you use it enough then it will also be in cache so the CPU does not have to access main memory which is on another part of the motherboard. This speeds things up significantly.
More information: https://www.youtube.com/watch?v=WDIkqP4JbkE and https://www.youtube.com/watch?v=rX0ItVEVjHc

@Labeeb: Are you seriously ready to pay 500$ for a graphics card? You can get 2 decent laptops for that price.

Show me a laptop that costs 250$ and is decent.
500$ is what I paid for my graphics card (and I bought two of them so 1000$ in total), and (depending on where you live, as prices vary dramatically from country to country) 500$ is a reasonable price if you want a powerful PC.

NVidia GeForce GTX980 is really good, but do you REALLY need that much power? Between the GTX980 and GTX970, the difference isn't particularly huge, and I get the impression that you are not planning on running more than 1 GPU. Are you willing to spend an extra $200 just for a slight improvement that you probably won't be using?

Btw, I am personally running i5 with one NVidia GeForce GTX760 (I bought it before the better 900 series came out), and I can already run pretty much any game just fine.

This is exactly what I would recommend.
Honestly for gaming an i7 CPU is not worth it as hyper-threading simply does not mean much in a game.
So unless you plan to be doing things like hardcore video editing then just settle for an i5 CPU, and instead spend the money you saved on a powerful graphics card.