Chronicle of development of auxiliary power supply interface of graphics card (brief history)

Graphics card With auxiliary power supply, it has become a matter of course, but its first appearance can be traced back to the age of AGP interface. Earlier PCI graphics cards can only display 2D graphics, and playing games basically depends on CPU computing. However, with the birth of games such as Tomb Raider in the early era, graphics cards had certain 3D graphics computing functions. Later, graphics cards accelerated to 3D in an all-round way OpenGL and DirectX With the birth and development of graphics API, graphics cards finally took on the task of graphics computing independently, greatly reducing the burden of CPU.

However, the 3D performance of the graphics card at the beginning of AGP is not high, so the 25w power supply capability of AGP interface is not under pressure. Although the GPU performance is not good at this time, it can be added in quantity, so the Voodoo5 5500 graphics card of 3dfx Company (acquired by NVIDIA) is equipped with two VSA-100 graphics cores, which are connected using SLI technology (you are not mistaken), and the overall power reaches 36W, so the power supply of AGP interface cannot be held, So for the first time, the video card is equipped with an auxiliary power supply interface. Pay attention to the white part on the upper right corner of the video card.

Since the acquisition of 3dfx, the 3D graphics card market has formed NVIDIA and ATI (which have been AMD Acquisition). However, for quite some time since then, the video card has never been equipped with auxiliary power supply. Even NVIDIA's high-performance GeForce4 Ti series video cards have not been equipped with auxiliary power supply until NVIDIA's GeForce5000 series and ATI's 9000 series are listed, such as GeForce FX5200 Ultra The GeForce FX5800 and FX5900 Ultra are equipped with D-type auxiliary power supply, also in the upper right corner.

The ATi was not equipped with an auxiliary power supply until the Radeon 9500, but it was a small 4Pin interface, that is, the power cable for the 3.5-inch floppy drive. However, this interface was obviously lacking in firmness, so later high-end models such as the Radeon 9700 PRO and Radeon 9800 began to be equipped with D-type auxiliary power supply interfaces. The following figure shows the Radeon 9500 graphics card. Also pay attention to the white 4Pin interface in the upper right corner.

So, what if the performance of the graphics card is too strong, and one D-type auxiliary power supply is not enough? The answer is to add another one. For example, the NVIDIA GeForce 6800 Ultra graphics card in the figure below is equipped with two D-type interfaces on the right side. Since then, the auxiliary power supply of the graphics card in the AGP era has come to an end, because it will enter the PCI Express era.

In the era of PCI Express, the power supply capacity of the graphics card slot has been improved to 75W, which can meet the power supply needs of mainstream graphics cards. However, for ultra high-performance graphics cards, auxiliary power supply is still needed, but the interface of auxiliary power supply needs to be changed for new cards. So there are the following two kinds of 6Pin and 8Pin, which are very familiar to everyone. The difference between the two is the power supply capacity. The power supply of 6pin is 75w, 8pin is 150w, 6+8pin is 225W, double 8pin is 300w, and three 8pins are 450W.  

Now the auxiliary power supply of the graphics card has evolved to the ATX3.0 era, and the interface of the graphics card has also changed into a 12Pin+4Pin interface, called 12VHPWR, and first appeared in NVIDIA GeForce RTX On 3090, 12Pin is responsible for power supply, while the smaller 4Pin is responsible for communication with the power supply. The actual output power can be adjusted according to the power of the graphics card. The maximum output power can be up to 600w, and even can cope with the instantaneous power of "Kahuang" such as NVIDIA GeForce RTX 4090.

At present, 12VHPWR has been applied to RTX 4060, RTX 4060 Ti and RTX 40 graphics cards of the upcoming RTX 4050. Intel also said that this interface will be used by future graphics cards, while AMD still uses 8Pin auxiliary power supply throughout the line. Do you think the next generation will use 12VHPWR interface instead?

This article is an original article. If it is reproduced, please indicate the source: Chronicle of development of auxiliary power supply interface of graphics card (brief history) https://diy.zol.com.cn/831/8311320.html