The difference between gpgpu and gpu: GPGPU is responsible for the calculation of some non-graphics related programs, while GPU is responsible for graphics rendering. The tasks of the two are different. 2. GPGPU is usually integrated into the CPU, while the GPU is used as a separate module, that is, an independent graphics core or a motherboard integrated graphics core. 3. The GPU is responsible for some graphics rendering and graphics calculation scenarios, and the performance of the GPU is crucial when playing games; while the GPGPU is responsible for some calculations that have nothing to do with graphics, which is specifically reflected in the speed of encryption and decryption, etc.
The operating environment of this tutorial: Windows 7 system, Dell G3 computer.
What is GPU
Currently, the computing requirements related to graphics processing are increasing, which is limited by the floating point computing power of the CPU itself. For the needs For high-density computing image processing operations, the traditional method of implementing it on the CPU has not made much progress in processing performance and efficiency. Therefore, the industry has developed GPU (Graphics Processing Unit) specifically for the computing needs related to graphics processing, that is, a microprocessor that completes image operations.
What is GPGPU
#With the rapid development of artificial intelligence technology, in addition to the demand for graphics-related computing power, there are also requirements for the general computing power of the CPU It's getting higher and higher. At this time, GPGPU applications were born.
The full name of GPGPU is General Purpose GPU, which is a general computing graphics processor. The first "GP" represents general purpose (GeneralPurpose), and the second "GP" represents graphics processing (GraphicProcess). The combination of these two "GPs" means "general graphics processing". GPGPU can be generally understood as a tool that assists the CPU. It can help the CPU perform calculations of non-graphics-related programs.
When designing the GPGPU architecture, the acceleration hardware unit designed by the GPU for graphics processing was removed, and the SIMT architecture and general computing unit of the GPU were retained. Therefore, current GPU-based graphics tasks cannot be run directly on GPGPU (it may be possible in the future), but general computing tasks such as scientific computing, AI training, and inference tasks (actually mainly matrix operations) still retain the advantages of GPU, that is, Efficiently transport, compute, and repeat tasks involving massive amounts of data. Currently it is mainly used for physical computing, encryption and decryption, scientific computing, and the generation of cryptocurrencies such as Bitcoin.
The difference between gpgpu and gpu
1. Different tasks
GPGPU is responsible for some The operations of non-graphics related programs, and the GPU we often hear is responsible for graphics rendering, the tasks of the two are different.
2. Different forms of existence
GPGPU is usually integrated into the CPU, while the GPU is used as a separate module, which is an independent graphics core or motherboard Integrated graphics core.
3. Different functions
Since the two are responsible for different areas, their functions are naturally different, which is reflected in daily use: GPU is responsible In some graphics rendering and graphics computing scenarios, the performance of the GPU is crucial when playing games; while the GPGPU is responsible for some calculations that have nothing to do with graphics, which is specifically reflected in the speed of encryption and decryption, etc.
Extended knowledge:
In order to better distinguish the difference between GPU and GPGPU, AIDA64 can view GPU-related information:
Select "Display Device" - "Graphics Processor (GPU)" in the left menu
Click to view GPU related information , which includes GPU process, manufacturer and other information.
Use AIDA64 to view the detailed information of the GPU
For more related knowledge, please visit the FAQ column!
The above is the detailed content of What is the difference between gpgpu and gpu. For more information, please follow other related articles on the PHP Chinese website!