Shulou(Shulou.com)06/03 Report--

What is GPU?

First, you need to explain what the abbreviations CPU and GPU stand for, respectively. CPU is the central processing unit and GPU is the graphics processing unit. Second, to explain the difference between the two, we must first understand what they have in common: both have bus and external connections, have their own cache system, and digital and logical computing units. In a word, both are designed to complete computing tasks.

The difference between the two lies in the structure of the cache system and digital logic operation units that exist on the chip: although CPU has multiple cores, the total number does not exceed three digits, each core has a large enough cache and enough digital and logical operation units, and assists the hardware with many accelerated branch judgments and even more complex logic judgments; the core number of GPU far exceeds that of CPU, which is called multicore (NVIDIA Fermi has 4 cores). Each core has a relatively small cache size, and digital logic operation units are small and simple (GPU has always been weaker than CPU in floating-point computation at the beginning).

As a result, CPU is good at dealing with computing tasks with complex computing steps and complex data dependencies, such as distributed computing, data compression, artificial intelligence, physical simulation, and many other computing tasks. Due to historical reasons, GPU is produced for video games (so far its main driving force is the growing video game market). A kind of operation that often occurs in 3D games is to perform the same operation on massive data, such as carrying out the same coordinate transformation for each vertex and calculating the color value for each vertex according to the same lighting model.

GPU's multi-core architecture is very suitable for sending the same instruction stream to multiple cores in parallel, using different input data to execute. Around 2003-2004, experts in fields other than graphics began to notice GPU's distinctive computing power and began to try to use GPU for general-purpose computing (that is, GPGPU).

GPU accelerated computing refers to the simultaneous use of graphics processing units (GPU) and CPU to accelerate the speed of scientific, analytical, design, consumer, and enterprise applications. The GPU Accelerator was first launched by NVIDIA in 2007 and now supports energy-efficient data centers in government laboratories, universities, companies, and small and medium-sized enterprises around the world. GPU can accelerate applications from cars, mobile phones and tablets to platforms such as drones and robots.

In summary, the purpose of GPU is as follows:

GPU was first used to improve the performance of graphics computing and improve the quality of graphics.

Nowadays, GPU is not limited to graphics processing, but also can be used to accelerate parallel computing.

Computer-aided engineering (CAE) simulations allow engineers to design more virtual prototypes and therefore spend less time building physical prototypes. More repeated designs can lead to higher quality products. By accelerating simulation, GPU can further improve productivity, which helps to shorten product development time and bring more competitive advantages for enterprises.

ANSYS works closely with NVIDIA to ensure that simulations running on ANSYS parallel software achieve performance. NVIDIA GPU supports the following products:

Structure: ANSYS Mechanical 17.0 supports a full set of features, including running on multiple GPU. Fluid: ANSYS Fluent 17.0 provides GPU support for pressure-based coupled solvers and radiative heat transfer models. Electromagnetic: ANSYS HFSS 17.0supports instantaneous solver for transient flow simulation.

It is now easier than ever to accelerate ANSYS simulations with NVIDIA GPU. In version 17.0, all HPC licensed products (HPC, HPC Pack, HPC Workgroup) support GPU. Specifically, in terms of license requirements, each GPU is regarded as a CPU core. In this way, using the existing HPC license plus NVIDIA GPU, you can greatly increase the simulation productivity, thus enabling more simulation work to be done.

High performance Computing (HPC&GPU) is the development trend of simulation computing in the future, and it will be one of the important signs of national competitiveness.

As the driving force of high-performance computing, GPU is a perfect combination of high-performance and low-power consumption.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.