Skip to content
【☢】 BitHardware.top ▷ Hardware, Reviews, News, Tutorials, Help post

Why you shouldn’t use a professional GPU for mining

If you have a GPU, you will have considered being able to mine with it, whatever the type. However, the price increase does not seem to have affected the so-called professional GPUs, which are just as good as gaming for mining. However, there is an explanation why this phenomenon has occurred and that is directly related to the mining capacity of these GPUs.

What elements affect cryptocurrency mining?

There are three factors that we have to take into account, which are the following:

  • First of all, the mining rate or the speed at which the GPU performs the function. The more computing power and speed in memory a GPU has, the higher its mining rate.
  • Second, that your mining ratio is balanced with your selling price. There may be a less powerful GPU, but at a much lower price than much higher benefits.
  • The third and last is the cost of the infrastructure, which includes both the hardware that is necessary to mine as well as the energy cost. Regarding the cost of infrastructure, we will focus on it as an argument in this article, energy depends on the cost of electricity in the country or area of ​​the world in which we are, as well as taxes on electricity.

Once we have remembered these three points, let’s go to the explanation.

Professional GPUs vs gaming cards

Quadro_RTX_A6000

A GPU for gaming, such as NVIDIA GeForce or AMD Radeon, is a means of consumption. At the end of the day we do not use it for work, although there are exceptions and therefore it is not a tool.

On the other hand, a professional GPU is used as a means of production, since it is a working tool in certain sectors. Although at the architectural level they are almost similar to each other, professional GPUs have an associated ID that allows them to have higher quality graphics drivers, automatic color calibration systems or even support for virtualization, which allows their installation on servers. in the cloud or for use with various operating systems.

From a certain point of view we can say that all hardware can be transformed into a means of production, but it is the level of capitalization of the same that gives it its value. And this is what has affected gaming GPUs with the emergence of the mining boom. Which has turned them into a means of production, since mining turns them into an indirect way of getting money. So its value has risen enormously, to stratospheric levels compared to its recommended sale price.

Why are professional GPUs not good for mining?

NVIDIA Quadro

The answer is that it is a lie that professional GPUs are not suitable for mining. The correct answer is that its price is so high that it takes much longer to raise money with similar hardware in a professional GPU than using its gaming equivalent. It should be borne in mind that although they have much more memory, they do not increase their total bandwidth, but rather use VRAM’s so-called clamshell modes to place twice as many memory chips per data interface. Thus doubling the capacity, but not the bandwidth.

Another reason, and in the exclusive case of NVIDIA, can be seen by observing the specifications of its professional A6000 GPU compared to the RTX 3090. While the gaming GPU has 24 GB of GDDR6X memory, the professional has a configuration of 48 GB, but with a slower GDDR6. Which negatively affects performance when mining, since some algorithms such as the one used in the Ethereum cryptocurrency are highly dependent on bandwidth.

But in general it is the high price of professional GPUs that causes them not to be used for mining. Not because of technical limitations and the GPUs they have installed are the same as gaming ones, but because mining to be profitable has to pay the acquisition costs of this type of card.

And what about accelerator cards or just computing?

NVIDIA-Ampere-GA100-GPU_7nm-Tesla-A100

GPUs today are not only used to generate graphics, but also in all kinds of disciplines to crush numbers in large quantities. All this thanks to its enormous capacity to execute algorithms in parallel at high speed, which has led to any supercomputer not being based on a large number of CPUs, but also on GPUs integrated into the system.

Mining is not a discipline based on the execution of graphics, so it is a task for computing. So it would be ideal in the case of those GPUs designed for that market. Which are usually sold to large universities around the world, in defense contracts, large pharmaceutical companies, and so on. So we are talking about hardware that is sold with large economic contracts and away from the ordinary user, even if we talk about professionals.

However, we have cases like the NVIDIA Tesla in its PCI Express version that can be purchased for a “modest” of more than 10,000 euros, as can be seen above. Its performance is clearly better than that of the RTX 3090 and A6000, but its high price makes it totally impossible to monetize your purchase. In other words, they are highly inefficient from an economic point of view when it comes to mining. Hence, as with professional GPUs, they have not seen their sale price in the clouds. A proof that they are not efficient is in the fact that many of the companies that own this type of systems equipped with these GPUs have not put their supercomputers and data centers to mine.