So I know CPUs have different clock speeds etc. but what exactly makes a CPU faster or better?
Why is e.g. my 5 year old 4-core 3GHz CPU worse than my new 4-/6-core 3GHz one?
In: Technology
In addition to new and more hardware like L1, L2 cache, smaller cores meaning shorter pathways, the technology inside has been improving as well. Couple generations ago, the cores started adding predictive analysis for the logics. They would try to predictive which operations were going to be done next based on history and setting those channel up ahead of time. Rather than solely reacting to input command in previous generations.
It pretty much all in how it’s built, and what architecture it’s built on. The better it’s built, the more power efficient it would be, and the more powerful it may be.
There is more to a CPU than just clock speed such as the size and type of cache, bus type, whether or not it has hyperthreading, amount of cores, thermal output, etc.
A CPU with 4 cores and no hyperthreading on an older bus type will perform poorly compared to a newer CPU with 6 cores that has hyperthreading with the latest bus type that optimizes for speed when they have comparable clock speeds.
[here](https://www.pugetsystems.com/labs/articles/Specs-Explained-CPU-137/) is a decent article explaining what different CPU specs actually mean as your question itself is flawed and incapable of being answered.
The clock speed of a CPU is like a drummer when an army is marching. It keeps everything synchronized.
The clock insures no signal gets sent, before something is ready to receive it.
CPUs starting with the Pentium started using something called pipelining with the Pentium series. Not every operation could be completed in a single cycle, but by using a pipeline one instruction could be started on one cycle, and the next instruction started on the next cycle.
Problems occur when the next instruction is dependent on the result of the previous instruction. You can either wait for the results on the other end of the pipeline, or make a guess what the result will be. AMD Athlon processors had a shorter pipeline than Pentium 4s. That meant that Athlons with lower clock speed were comparable to P4s with higher clock speed.
Now, this type of efficiency is called IPC: Instructions Per Cycle
The performance of modern CPUs is all about IPC.
Having more cores allows more independent threads to be executed at the same time, increasing IPC.
Having more cache memory on the chip means the cores can spend less time waiting on main memory when dealing with large data sets, (often one thread will pull the needed data into the cache, while another processes the data, since the cache is accessible to all the cores on the die.)
Faster memory interfaces also get data in and out of cache faster, so the CPU will have fewer idle cycles where all it is able to do is wait.
In short: we’ve kinda gotten to the limits of how fast we can make the components in a CPU “march” but we’ve found ways to make individual components smaller so we can have more of them in a limited space. By having more they can work in parallel.
Latest Answers