Building your own gaming PC ? perhaps you want to play with more than 30FPS, welcome aboard the PC master race, or maybe just curious to know what are GPUs, you have hit a jackpot so let the geek in you rise.
What is a GPU ?
A GPU or Graphic Processing Unit is a chip which has way more cores compared to CPUs and does huge data calculation. Even the screen you are viewing this article on is using a but of GPU processing. A GPU is nowadays necessary to even run your Operating System smoothly, whereas in mobiles 3D games are becoming more engaging and graphically enhanced than ever before and to run those games smoothly you need a powerful GPU.
A GPUs performance is measured in FPS or Frames Per Second , that is how many frames can the GPU render in one second, 60+ FPS means is a lag free game.
Types of GPUs
GPU category is broadly divided into two categories , Dedicated GPUs and Integrated GPUs.
Dedicated GPUs are often more powerful and can cost you extra $_$ but they are also much more capable compared to their counterpart. A dedicated GPU is present in almost all smartphones being manufactured nowadays. While dedicated GPUs in Computers are way more powerful and often require their own separate power, and a PCIe x16 slot.
Integrated GPUs are mostly famous in the PC world and are useful for a person who does not depend a lot on heavy 3-D related stuff, a integrated GPU is a more economical and a viable option for many, although not as powerful as a dedicated one nowadays a integrated GPU can do way more tasks other than basic stuff it can also run light 2-D/3-D games, A integrated GPU is inbuilt in a CPU which serves as a CPU and also for some graphical use.
Why is the difference between CPU and a GPU ?
Well a CPU has more of a general purpose use, it can do many things like light gaming, operating your system, making presentations, think of it like a multi purpose tool which has a knife scissors, bottle opener etc, now GPUs are very specific in nature, their job is to take in raw input and perform millions of calculations based on physics and render the output, so here a GPU can be described as a surgical knife now a doctor would not use the multipurpose knife over the surgical knife will he? even though multipurpose knife will cut through but the surgical knife does a better job at it as it is made to do just that, A GPU has way more sub cores than a CPU, a CPU might have at most 8 very powerful cores whereas a GPU has around 1000s much smaller sub cores however a CPU core is more diverse at what it needs to do and also way more fast compared to the GPU cores they do much complex calculation compared to what a GPU does, also GPU cores have to be dedicated to solving a parallel problem and GPU mostly solve simpler problems like geometry calculations which if presented to small number of CPU cores will clog up the cores as 100 calculation going through 8 cores will take time, and if 100 questions is given to 1000s different sub cores is a more efficient and quicker.
12GB GPU vs a 8GB GPU is there a big difference ?
Lets understand what the use of VRAM is on a GPU, we have RAM in your system that holds information to be fed to CPU for processing similarly we have VRAM (Virtual RAM) in GPUs which store information like textures Frame Buffer lighting information etc as it is much faster for the GPU to get information from the VRAM rather than the SSD or your HDD to keep with the output. Things like your resolution and anti-aliasing affect (the jaggy edges which get smother) the most, for example if you are running a multi-monitor setup or maybe you are running a game on 4K instead of the FHD resolution, that were you need more VRAM. So is there a big difference ? well if you have more VRAM than what your system is using then the left VRAM is just a waste as it wont help boost your Frames per second on the flip side if you have less VRAM there is a huge problem as you will feel texture popping shuddering and a poor performance therefore it is best to know how much do you require and research quite a lot before going for your GPU.