Different types of co-processors

               TYPES OF CO-PROCESSORS

                       

we know that CPU(central processing unit) is heart and brain of the computer which used to perform mathematical calculations ,execute the instructions. CPU is versatile, It can perform n number of operations but their is a situation where CPU has to perform more number of tasks. it is time consuming and cause inefficiency the only solution is share the work load to different types of co processors like graphical  processing unit(GPU)  tensor processing unit(TPU)  artificial intelligence processing unit (APU)  neural processing unit(NPU) holographic processing unit(HPU) in this article we try to see these processors in detail

GPU (graphical processing unit)

        

   

    If we are playing high end game in mobile phone or pc's we Definitely need the graphic processing unit (GPU) it is used for  operating system interface ,graphics ,rendering 3d Modeling ,animations ,video editing .we know that there are millions of pixels on screen if we are playing a 40fps(frame's per second) video we have to do millions of calculations to refresh the pixels to represent on the screen. it hard to manage this type of calculations by the  CPU and it takes more time ,so the GPU is very important can be available  in integrated or dedicated GPU

TPU (Tensor processing unit)

                    

It is the most advanced processor that developed by the google. It is specially design to accelerate machine learning workload. every co-processors have their own processing tasks and also used together. not only google other company's are also working to build their own processors for computing  machine learning algorithms for example google lens app in our mobile process the image in real time this is because of the machine learning processor units in their servers.

NPU (Neural processing unit)

                    

This neural network processing unit work similar to our brain. It have parent and child node this NPU have the ability to learn  itself by giving inputs to this unit. And it is specially design to compute the task that are related to neural networks. This type processors we can see in Kirin and also apple company developed neural engine  

APU(artificial intelligence processing unit)

                 

   Generally we unlock our phone with face recognition this because of APU unit to unlock our phone the APU have to perform millions of tasks . Media Tech  company use this type of chips in their product  so it work more efficient and fast also it used to perform many tasks that are related to the APU

HPU(holographic processing unit)

                           

   holographic processing unit have very wide range of applications it is used in constructions 

 multimedia ,medical ,VFX ,military mapping ,arts etc. soon this type of HPU will come in mobile phone and we experience in real life 

 

there are some other co-processors that are used in servers and satellite rockets etc..

this type of co-processors are used unless we use it ..for example if we want to render a video for better editing the programmer have to program the CPU that this type of work load  transfer to GPU similarly HPU,NPU,APU  there is no use of co-processors until or unless we use it right..!!!