For those who work with cryptomining, the use of processors to mine is very important. Thus, knowing the basics about processors is necessary to start in the world of mining. This is why we have decided to briefly explain the two processors that are used for this: CPU and GPU. We hope it helps
A CPU or Central Processing Unit is a processor designed to perform a sequence of instructions. Thus, this has as main objective the loading of programs. According to the architecture of von Neumann (architecture with which the CPUs work), it’s divided into 4 steps: reading, decoding, execution and writing. Thus, it’s a multipurpose processor, since it can perform a number of tasks at the same time, regardless of the complexity and variety of tasks.
Thus, the processor retrieves the instructions from the memory, decodes to interpret the instruction, performs the operations necessary to execute said instruction and writes the previous results.
The GPU or Graphic Processing Unit is a coprocessor that is dedicated to floating point operations. This coprocessor is designed to reduce the workload of the processor or CPU. The GPU has a high segmentation, being that it consists of a large number of graphics cores, or «small CPUs» as many explain, with very specific purposes.
Thus, it can be considered dependent on the CPU, since it receives the same information through vertices. Its memory is faster and it allows managing intermediate results of mathematical operations in a much more agile way. Now, mining with GPU is the process of mining through a graphics card or GPU. Performing mining with this processing unit allows generating a great power of floating point operations per second (FLOPS).
Being that both are known processors and are used to mine, as well for the common processing of computers, knowing their differences is important to know which may be the best for the work they need.
The first of these differences is the speed of processing each one. Being that the GPU is much more specialized than a CPU processor, and its thousands of cores work in parallel, multiply by 100 the performance of a computer, when we talk about operations with vectors and matrices. Thus, its speed is greater than that of the CPU, which is more multipurpose than the GPU.
Another of the main differences is the specialization of each processor. The GPU is a specialized processor, so it performs certain tasks. Work with large amounts of data and perform the same operations. However, these tasks will not be done quickly if they are very complex. Meanwhile, the CPU is multipurpose, so it can perform a small amount of tasks, but these can be very variable and complex than those that a GPU can perform.
Finally, as we said, the GPU can be dependent on the information vertices of a CPU. Thus, it is necessary to use a CPU to have a GPU. While with a CPU processor no other processor is required to function properly.