igpu heterogeneous capabilities: 1. It is heterogeneous computing, allowing the increasingly powerful converged GPU to assist the CPU in computing through heterogeneous computing technology; 2. Compared with serial computing, parallel computing can be divided into time Parallelism and spatial parallelism; 3. Minimize the execution time of computing tasks.
The operating environment of this tutorial: Windows 7 system, Dell G3 computer.
igpu heterogeneous capabilities:
1. It is heterogeneous computing, allowing the increasingly powerful converged GPU to assist the CPU in computing through heterogeneous computing technology, that is to say The proportion of GPU calculations is increasing, and software is more dependent on GPU operation. For example, if you use DX12 software that is often used for playing games, the performance gap between using DX12 and not using it can quickly double.
2. Compared with serial computing, parallel computing can be divided into time parallelism and space parallelism. Temporal parallelism is pipeline technology, and spatial parallelism uses multiple processors to perform concurrent calculations. The current research is mainly on spatial parallelism. From the perspective of program and algorithm designers, parallel computing can be divided into data parallelism and task parallelism. Data parallelism breaks large tasks into several identical subtasks, which is simpler to process than task parallelism.
3. The ultimate goal it pursues is to minimize the execution time of computing tasks.
Extended information:
(1) Heterogeneous computing technology emerged in the mid-1980s because it can cost-effectively acquire high-performance computing capabilities and has good scalability. , high utilization rate of computing resources and huge development potential, it has become one of the research hotspots in the field of parallel/distributed computing. The authors of this article also highlight some aspects that should be paid attention to in future heterogeneous computing research.
(2) With the rapid development of communication and network technology, the concept of network computing emerged as the times require. Homogeneous network computing systems now or cow first emerged, followed quickly by heterogeneous network computing systems, making heterogeneous computing one of the main research hotspots in the field of parallel/distributed computing in recent years.
(3) Parallel computing performed on heterogeneous computing systems is usually called heterogeneous computing. People have defined heterogeneous computing from different angles. To sum up, we give the following definition: Heterogeneous computing is a special form of parallel and distributed computing. It may use a single independent computer that can support both SIMD and MIMD methods. Computers, or a group of independent computers interconnected by a high-speed network to complete computing tasks. It can coordinately use machines with different performance and architecture to meet different computing needs, and enable code (or code segments) to be executed in a way that maximizes overall performance.
(4) The computing resources it uses have various types of computing capabilities, such as simd, mimd, vector, scalar, dedicated, etc.;
(5) It needs to identify the computing tasks The type of parallelism requirements of each sub-task;
(6) It needs to enable computing resources with different computing types to coordinate with each other;
(7) It requires both the development of application problems and Parallelism, but also the development of heterogeneity in application problems, that is, the pursuit of matching between the computing type of the computing resource and the type of task (or sub-task) it performs;
Related free learning recommendations:php programming(video)
The above is the detailed content of What is the heterogeneous capability of igpu. For more information, please follow other related articles on the PHP Chinese website!