The Central Processing Unit, or CPU, is the silicon chip that handles the majority of all operations on your computer. Backed by billions of microscopic transistors, it translates instructions from software into the electric signals that execute those tasks.
Two major specifications used to compare processors are core count and clock speed. The former refers to how many physical processor cores a chip has while the latter is the number of cycles per second it can perform. Core counts can be misleading, however, as it is the overall performance of those cores that matters.
A CPU works best for serial computing processes like arithmetic and logic, turning data input into more usable information output. For example, the natural language processing (NLP) algorithms used by robots and smart home devices, or Markov models and support vector machines for drones and autonomous vehicles run well on CPUs.
When an instruction is fetched, it goes through an instruction decoder which converts it into a series of electrical signals that are sent to different parts of the CPU for execution. These could include adding numbers, performing logical functions like Boolean logic, comparing numbers, or jumping to a RAM address. After an instruction is executed, the program counter is incremented and the next instruction is fetched.
A Graphics Processing Unit (GPU) on the other hand, is a more specialized component that excels at highly repetitive, massively parallel computing tasks. The GPU is able to render every single pixel of computer graphics on your screen far faster than the CPU can, thanks to its large number of cores. Cpu’s