Theoretical part
Recently, Nvidia introduced the new GeForce RTX 30 family of graphics cards based on the Ampere architecture, which replaced Turing. The previous architecture was revolutionary, for the first time offering hardware support for ray tracing and hardware acceleration of artificial intelligence tasks using tensor cores. But the performance of those GPUs was sometimes not enough even to use a couple of effects using tracing, so it’s not surprising that Nvidia focused on performance in Ampere.
If we compare video cards, then the GeForce RTX 3090 is not closer to the RTX 2080, but something in between the RTX 2080 Ti and Titan RTX, as they are the top representatives of their generations.
Architectural features
The GA102 uses Samsung’s 8nm process technology and is further optimized specifically for Nvidia. The older Ampere gaming chip contains 28.3 billion transistors and has an area of 628.4 mm² – this is a good step up from Turing’s 12nm, but the same 7nm process at TSMC still surpasses Samsung’s 8nm in density, judging by on chips of the same Ampere architecture, comparing the gaming GA102 and the large GA100 chip, which is produced in Taiwanese factories.
The GeForce RTX 3090 graphics card is basically the most suitable for 8K resolution. First, instead of using two cables to connect to HDMI 2.1 devices, only one cable can be used on all Amperes. Secondly, this is the most productive graphics processor today, and even 4K resolution requires such power, not to mention 8K. And thirdly, often high-resolution 3D applications use a lot of video memory, and the GeForce RTX 3090 has more than twice the amount of video memory compared to the company’s last flagship – RTX 2080 Ti with its 11 GB. The new model carries 24 GB of the new type of GDDR6X memory with high bandwidth, which is also important for 8K. So Nvidia is not too far from the truth when it calls this GPU the first suitable for 8K.
Application in professional software
The GeForce RTX 3090 graphics card is designed not only for wealthy PC gaming enthusiasts but also for various representatives of the modern industry who use the power of modern GPUs in their own work: designers, 3D animators, AI developers, scientists, and many others who aim to use the most powerful GPUs in the world. It is the GeForce RTX 3090 that is, offering the maximum of everything. Similar to its conditional predecessor in the form of Titan RTX, the novelty carries 24 GB of fast memory on board, which is important for all these categories of users, in particular for processing video data in 8K HDR format, and also offers the highest performance in computing applications, including those with hardware support. ray tracing.
The GeForce RTX 30 series is great for all of these applications, with new features including 4th generation PCI Express support that doubles the bandwidth between the GPU and the rest of the system, helping to transfer huge amounts of data when processing high-definition video. For the same purpose, a very large amount of local video memory of 24 GB will be useful. Many video applications will allow you to comfortably work with a variety of effects when processing 8K video. And in applications using ray tracing, the average performance of solutions based on the Ampere architecture is 1.8-2.4 times higher than that of similar GPUs from the Turing family:
Thus, we see that the Nvidia 3090 does an excellent job of all tasks, but is still too expensive for gamers. And what about mining? Can the flagship RTX 30 series be of interest to miners? To understand this, you need to look at the 3090 hashrate charts:
At the moment, this video card brings about $ 1.7 net on the ETH cryptocurrency and 1.3 on Conflux and Cortex. Good performance, despite the general collapse of the crypto market.
Summing up, we can recommend 3090 for purchase, it will be a worthy long-term investment, which will be relevant for many years to come.
