Quantcast
Channel: I. Axcess – INSCMagazine
Viewing all articles
Browse latest Browse all 75

How Does a Reference Video Card Differ From a Custom One?

$
0
0

There are a great variety of video cards on the market. Graphics chip manufacturers usually provide their own versions, called reference versions. But third-party vendors prefer to develop their own versions, which are called custom. What is the difference between reference and custom video cards? And what’s better? We tell you in our material.

Pro Tip: Asif Computers has a wide range available in the MacBook Air category if you are also looking to buy a MacBook Air for yourself then click on the MacBook Air price in Pakistan in our given link now.

Board and power subsystem

The video card board plays the same role for the graphics processor as the motherboard does for the computer’s central processor. It is on the board that the power subsystem (VRM) is located, which converts 12V from the power supply into the voltage required for the GPU.

The design of the reference board and VRM is usually created with an eye on the maximum capabilities of a particular graphics chip, and with a small margin. If there are several models of video cards on the same GPU, the same reference board is used for all of them. But with the difference that on weaker video cards the power subsystem can be simplified. For example, the GeForce RTX3070 in the reference design has nine GPU power phases, although there is room on the board for a tenth phase.

And the GeForce RTX3060 Ti, based on the same chip and board, has eight phases. In this case, not one, but two phases are simply not soldered, although there is still space for them on the board.

Custom boards can either match the capabilities of the reference ones or differ from them. For inexpensive models, simplified boards are usually produced. They are more compact in size than the reference ones, and initially contain fewer electronic components and seats for phases. Such boards cost manufacturers less. For example, the ZOTAC GeForce RTX3060 Ti Twin Edge has only eight places for phases on the board, and only six of them are soldered.

For top models, on the contrary, more advanced boards are being developed, which, compared to the reference ones, can be equipped with a larger number of phases and other electronics to provide power reserves during overclocking. For example, the MSI GeForce RTX3060 Ti Gaming X Trio has eight phases, but the board is designed for as many as eleven.

Often such boards are larger than the reference ones, although sometimes there are compact options. If PCI-E 6/8 pin additional power connectors are used, then boards for top models often have more of them than regular ones: instead of one there are two, and instead of two there can be as many as three.

It is worth noting that custom video cards do not always use their own boards. Some solutions are equipped with reference boards, and only the cooling system gets its own design.

Cooling system

The video card cooling system is the part that is most often customised.

Reference cooling systems for video cards are developed with an eye to the greatest efficiency in any type of case. Depending on the model of the video card, in such a system you can find the usual design with airflow from several fans, a radial fan (the so-called “turbine”), and even liquid cooling.

But among the latest generations of reference models, you can no longer find “turbines” or liquid cooling. Despite this, video card manufacturers are still developing their own cooling systems, which may have some advantages compared to the reference ones. Some models have larger and quieter ones than the reference, some have the opposite – comparable in noise, but more compact. Thanks to this, you can always choose a model to suit your case size and taste.

Different manufacturers use fans with a different number of blades and a specific cut, evaporation chambers and heat pipes. Combining all this in certain combinations, various implementations of cooling systems are obtained. For example, there are custom COs with a longitudinal arrangement of heat pipes, and there are others with a transverse arrangement. The technology of contact between the tubes and the graphics chip is also different: it can be either direct contact or an evaporation chamber at the base of the radiator.

Pro Tip: If you also want to buy a new good and cheap MacBook for yourself, then click on your given link Apple laptop price in Pakistan now.

Due to many targeted modifications, custom cooling systems are often quieter than reference ones. But it is not always more effective, especially in poorly ventilated buildings. Therefore, before purchasing any video card, you need to carefully familiarise yourself with the capabilities of its cooling system by looking for reviews with temperature data under load.

Specifications

In addition to the board and cooling system, vendors often make changes to the technical characteristics of the video card. Once upon a time they even used inactive units in graphics chips, but at the moment this is a thing of the past. Nowadays, frequencies are most often changed, and in some cases, models with non-standard video memory are also produced.

The reference core frequencies set by GPU manufacturers are currently not changed, probably only by the lazy. Almost all modern custom video cards have slightly increased frequencies, which allows you to declare overclocking and add the prefix OC (OverClock) to the model name.

In fact, there is very little sense in such overclocking. Typically, vendors add 10 to 200 MHz to the stock GPU frequency. However, modern chips operate at frequencies above 2 GHz, so this increase only provides a few percent of additional performance, which cannot be seen with the naked eye.

For example, the reference GeForce RTX4090 has a boost frequency of 2520 MHz, while the top-end Asus ROG Matrix RTX4090 Platinum has a boost frequency of 2670 MHz. Even this difference only adds up to about 5% additional performance, and other versions have even less. Top-end custom graphics cards overclock a little better, so you can get a couple of percent extra speed from them. But in any case, such a difference will be almost invisible.

Graphics memory in modern models is rarely overclocked, and usually only slightly. Therefore, we can say with confidence that now any factory overclocking of video cards is done more for show. But it is worth noting that this was not always the case. In the 2000s, you could find models that were much faster due to overclocking out of the box. For example, the 2008 budget card Gigabyte GeForce 9500GT was released in a version with core overclocking to 650 MHz, while the stock value was 550 MHz. The productivity gain in this case was about 18%.

Occasionally, vendors experiment with the amount of video memory. For example, the GeForce GTX750 comes with 1 or 2 GB of memory by default, and custom versions may have 4 GB.

It may also be possible to use different memory and graphics chips in the same video card model. But GPU manufacturers usually officially allow this feature.

Custom or reference?

In promotional materials, custom video cards are always extolled over reference ones: they say that they have lower temperatures and noise, and higher performance. This is often true, but not always. References with “turbines” are a thing of history, and their modern representatives are often equipped with quite sophisticated and relatively quiet cooling systems.

Therefore, it is possible that some custom models may be worse than the reference model both in terms of temperatures and noise produced, while some, on the contrary, are better. Before choosing a video card, in addition to price and dimensions, it is better to inquire in advance about the noise and temperature characteristics of the model you are interested in. Reference performance here is not a death sentence at all.

As you already understand, you don’t need to listen to “fairy tales” about increased performance either: it is practically at the same level for both reference and custom video cards. Except for a few additional percentages, which the eye is unlikely to notice anyway.

To summarise, we can say that decent models are now found among both reference and custom video cards. Therefore, you can safely choose the one that suits you in terms of size and noise, regardless of its performance.

For more Information: The inscribermag

The post How Does a Reference Video Card Differ From a Custom One? appeared first on INSCMagazine.


Viewing all articles
Browse latest Browse all 75

Trending Articles