Graphics Card of Gaming PC


 iStripper affiliate  iStripper affiliate

 To the Beginning  |  1  |  2  |  3  |  Towards the End

 nVIDIA logo At first, every PC had on their motherboards simple CPU and GPU devices (chips). In time, both devices are construction progressed and the CPU cores got math coprocessor (as shown in Figure 3.5.3), and graphics card with text mode began slowly to support elementary graphics functions. Common to them is whether they are also able to make only one PROCES, which is dependent on the capabilities of the operating system. Over time, software (especially games) has become more demanding and increasing the number of cores in the CPU and GPU to simultaneously could do more processes. The following Figure shows the composition of modern CPU and GPU. A larger number of cores in both cases has enabled by improved lithographic process of making chips.

 CPU vs GPU  iStripper
Figure 2.1 Core organization of CPU and GPU.  

The designs of CPU and GPU are significantly different by number of cores, which make the way of executing the instructions totally different. The CPU consist of four, eight or more cores, but the GPU has hundreds or thousands of cores. That makes the GPU able to execute hundreds or thousands of threads or processes in parallel by CUDA. CUDA is acronym of 'Computing Unified Device Architecture', it is an extension of the ' C , C++ ' programming language developed and introduced by nVIDIA in 2006. Using CUDA programming modules allows the programmer to take advantage of the massive parallel computing power of nVIDIA graphic cards, in order to use it for general purpose computation. With this programming modules the programmer will divide the program code into two parts, the first part of the code will be executed on the CPU while the second part of the code will be executed on the GPU. As programmer can use a CUDA-enabled GPU for general purpose processing - an approach termed GPGPU (General-Purpose computing on Graphics Processing Units). Basically, the CUDA platform is a software layer that gives direct access to the GPU's virtual instruction set with parallel computational elements, and their execution at multitude of kernels.

As a graphics card is installed into a PC, will see only connectors at the back side of PC. Many graphics cards have multiple outputs, so more than one display device can be used at a time. There are many kinds of graphic card outputs, interfaces, graphic processors and technologies. Let describe part of that.

Description of the graphics card was downloaded from the site 'Tom's Hardware' (, section /reviews/graphics-beginners,1288.html), by author Don Woligroski, created in 2006 on multiple pages. Here is a smaller part of the entire content, and if you are interested in more detail, I recommend visiting the mentioned site. A very interesting site is 'Gamesear' (, compatible with mobile devices and protected by SSL features, that deals with description and review of games. It is worth visiting, especially in the field of news.

As in any case, it is apparent from the description that the graphics card (GPU) performs a very complex job, and it is no surprise that the number of transistor units is more than the CPU. Graphics card is one of the biggest energy consumers in the computer and is probably one of the largest devices. And the price is part of the expensive component. To conclude, today's graphics cards using these technologies give very realistic views, and it's no surprise that more and more people love to play, including me.

 Why does the CPU have a lot less core than the GPU? A CPU core has to handle each single operation a computer does, calculation, memory fetching, IO, interrupts, therefore it has a huge complex instruction set, and to optimize the speed of fetching instruction branch prediction is used. Also it has a big cache and fast clock rate. To implement the instruction set you need more logic thus more transistors more cost per core compared to the GPU. The GPU cores have less cache memory, simpler instruction and less clock rate per core, however they are optimized to do more calculation as a group. The simple instructions set, the less cache memory makes them less expensive per core.

Traditional GPU designs use a single geometry engine to perform tessellation. This approach is analogous to early GPU designs which used a single pixel pipeline to perform pixel shading. Today, The Streaming Multiprocessor (SM) group is the heart of the GPU. It performs vital functions such as pixel shading, tessellation, and physics and compute calculations. SM is highly parallel processor employing superscalar execution for optimal performance. Superscalar execution is a technique that allows sequential instrutions from a program to be executed in parallel. Unlike thread level parallelism which improves throughput, superscalar execution also improves latency since the same program executes in less time.

At Figure 2.1 is shown one group of SM. In GPU can be implemented much more groups which makes GPU device more powerful. Beside that, NVIDIA provides a complete toolkit for programming the CUDA architecture that includes the compiler, debugger, profiler, libraries and other information developers need to deliver production quality products that use the CUDA architecture.

Nvidia also developed Scalable Link Interface (SLI) technology is or a multi-GPU technology for linking two or more video cards together to produce a single output. SLI is a parallel processing algorithm for computer graphics which allows for an increase in available data processing power. Therefore, it is not a problem to create a graphical server with multiple graphics cards associated with the way shown in Figure 2.3a. If multiple network cards are implemented, multiple such servers can be connected as shown in Figure 2.3b and get a rack with enormous graphical power.

 CUDA Server - Block diagram  
Figure* 2.3 CUDA server. ( + / - )  

Therefore, if one computer, tablet, cell phone, and the like, has a relatively weak graphics power, it can rely on the capabilities of the graphics server. A mode of operation that is a typical client-server architecture (Figure 3.6.3).


NOTE: It will continue!



 Girls  iStripper  iStripper  iStripper  iStripper  iStripper  iStripper




 To the Beginning  |  1  |  2  |  3  |  Towards the End

  Citing of this page:
Radic, Drago. " IT - Informatics Alphabet " Split-Croatia.
{Date of access}. <>.
Copyright © by Drago Radic. All rights reserved. | Disclaimer

 Home Page
 Extras  Informatics Alphabet