Vous êtes sur la page 1sur 20

It is a specialized electronic circuit designed to rapidly manipulate and alter memory in such a way so as to accelerate the building of images

in a frame buffer intended for output to a display. GPUs are used in embedded systems, mobile phones, personal computers, workstations, and game consoles.

Interest increased in early 90s with bitmapped graphics.

Games in mid 90s drove hardware 3D.


Late 90s, early 2000: we want flexibility!

The term was popularized by Nvidia in 1999. It marketed the GeForce 256 as "the world's first 'GPU', or Graphics Processing Unit, a single-chip processor with integrated transform capable of processing a minimum of 10 million polygons per second. Rival ATI Technologies coined the term visual processing unit or VPU with the release of the Radeon 9700 in 2002.

Modern GPUs use most of their transistors to do calculations related to 3D computer graphics. They were initially used to accelerate the memoryintensive work of texture mapping , later adding units to accelerate geometric calculations. GPUs include support from programmable shaders which can manipulate vertices and textures , oversampling and interpolation techniques to reduce aliasing, and very high-precision color spaces.

GPUs contain multiple cores that utilize hardware multithreading and SIMD (single input, multiple data). Having multiple cores and using SIMD processing means that GPUs have a lot of Arithmetic Logic Units(ALU).

GPUs utilize a wider SIMD width over CPUs, GPUs tend to have a width 32-64 compared to the SIMD width of 4 chosen by CPU designers.

Input:

Geometric model: Description of all object, surface and light source, geometry and transformations. Lightning model: Computational description of object and light properties, reflection. Synthetic view point ( or camera ): Eye position and viewing frustum.

Output: Colors/intensities suitable for frame buffer display.

3D models defined in their own coordinate system ( object space) Modeling transforms orient the models within a common coordinate frame (3D world frame)

Vertices lit (shaded) according to material properties , surface properties (normal) and light sources

Local lightning model (diffuse, ambient, etc.)

Maps world space to eye space Viewing position is transformed to origin and direction is oriented along some axis.

Transform to Normalized Device Coordinates (NDC) Portions of the object outside the view volume are removed.

The objects are projected to the 2D image place (screen space)

Rasterizes objects into pixels.


Interpolate values as we go (color, depth, etc.)

At this stage of the pipeline individual fragments (or pre-pixels) are assigned a color . The final colored pixels can then be displayed on a computer monitor or other display.

GPU cluster MATLAB acceleration using Parallel Computing Molecular modeling on GPU Bitcoin Mining

Recent developments in GPUs include support for programmable shaders which can manipulate vertices and textures, oversampling and interpolation techniques to reduce aliasing, and very highprecision color spaces. Moved from traditional fixed- function 3D pipeline towards a flexible General Purpose computational GPU.

GPU computing is a good choice for fine-grained dataparallel programs. The use of such hardware in a real-time system allow workloads to be supported that are too computationally intensive for CPU-only systems.

CPUs have few (two to eight) large, complex cores, while GPUs have up to a few hundred small, simple cores. They are focused solely on computation and complex assembly instructions.
GPUs are not right for every application. Only applications that have a substantial amount of parallelism can benefit from GPUs.

Graphics processing unit wikipedia NVIDIA -Gpu computing nvidia

GPU Architecture Overview- John Owen, UC, Davis NVIDIA. NVIDIA GeForce GTX 200 GPU Architectural Overview. May, 2008

Vous aimerez peut-être aussi