First of all we must attest to the Turing 12nm architecture that it was able to bring many new technologies to the world of graphics through the NVIDIA RTX 20 processor family. Among the most important of these technologies was RayTracing, which is an advanced way to make the gaming environment look like real life through simulation The rays of light track their path and the reflection of their effects and shadows on the atmosphere surrounding the game environment and the things inside it. This technology is very similar to GC technology used in cinematic films, and thanks to the DXR features that have been integrated into the Direct-X library, it has allowed developers and content creators the ability to display reflections and shadows within the gaming environment and produce vibrant graphic effects as if they were completely real life. Unfortunately, ray tracing technology is one of the challenges and obstacles facing game developers and players themselves due to its requirements that impose heavy amounts of complex data, which the RT Cores cores deal with with the GPU.
We also saw a remarkable progress in DLSS 2.0 technology, which is done through Tensor Cores cores and depends in its architecture on artificial intelligence for deep learning and high-performance computing to process low-quality distorted images and convert them into high-resolution images with more vivid and clear details along with accelerating the ability to process them and extract their results In faster time intervals. We also saw more new specifications in the RTX 20 family of processors, such as GDDR6, NVLink, DisplayPort 1.4, GPU Boost 4, and USB-C / VirtualLink VR, and don’t forget the better energy efficiency.
But we are now on the verge of the new Ampere 7nm architecture, and as you saw yourselves through the official Nvidia announcement of the RTX 30 series of cards, these cards have caused a boom in the graphics world due to the exceptional performance and reasonable prices for the upper and middle class. If you want our personal opinion, the new prices are already so competitive that they made the previous generation cards look like graphics cards from several generations ago, and unfortunately, these new prices will not play in the interest of users who own the current generation cards and intend to sell them at a loss to own the next generation cards.
But come with me, we look at the matter with a slightly different point of view than that of advertising technical sites. Despite the better performance of the next generation processors and some new technical specifications and features, it seems that the 7nm manufacturing accuracy used in the Ampere architecture did not come in its best form. Yes, the Ampere architecture brought the second generation of RT Cores, the third generation of Tensor Cores, and certainly for the first time the HDMI 2.1 port sees the light and we do not know what the RTX 30 cards have in the future, but did you not notice that we did not witness a noticeable change in the energy efficiency or the cooling system.
The strange thing is that the 7nm manufacturing accuracy is the same as the manufacturing precision used in the third generation AMD Ryzen 3000 processors and perhaps all of us have been evidence of AMD’s ability to break all energy efficiency barriers and achieve a revolutionary leap in performance levels from any previous architecture or manufacturing precision.
The new graphics cards “RTX 30” consume very high power rates, and this is certainly due to the huge amount of transistor graphics processors (we do not know their exact number yet) and the huge number of CUDA Cores. It is strange that the mid-range RTX 3070 card contains 5888 CUDA cores. This number with the cores of the RTX 2080 Ti card that contains 4352 cores, we will find a difference of 1536 cores, although it is only the next generation, it is expected that the RTX 3070 card alone will outperform the RTX 2080 Ti card in theory. So, in my personal opinion, if you are a class owner of high-performance graphics cards from the previous generation RTX 20, there is no need to change unless your budget allows and you have a 4K or 2K display with high refresh rates.
New cooling system, higher energy consumption
Next-generation cards will come with new requirements in energy consumption rates, the RTX 3070 card with the same requirements as the RTX 2080 Ti, perhaps the next generation cards would have come with lower energy consumption thanks to the 7nmFinFET precision manufacturing. The company’s only goal may be to face the current era’s graphics challenges and keep up with modern technologies in games through the largest amount of graphics processor resources and to confront the dense and complex data in this way. At the same time, this greater power will play a critical role in improving performance with large displays and enabling users to enjoy ray tracing technology and other graphic properties that require monsters in graphic processing.
The RTX 3090 will come with 350W TDP consumption while the RTX 3080 will come with 320W consumption and ultimately the RTX 3070 will have a 220W consumption rate and 93C maximum safe temperature. Thanks to a huge new cooler, the RTX 3090 takes up three PCI Express slots in the case. Also, some pictures appeared online that there is a new port in the RTX 3090 card which is a 12pin PCIe Connector, which we had no knowledge of except with this architecture, but at the same time the graphics card contains two 8Pin PCIe Power ports, and for this reason Nvidia recommends using the provider 750W power with each of the RTX 3090 and RTX 3080 cards, while the RTX 3070 comes with a rated power consumption of 220W and the recommended power supply is 650W, and you only need two PCI Express slots in the case. We’ll definitely see the wonders of cool and modular forms with next-generation graphics cards.
PCIe 4.0 data transfer interface
Let us not forget the fourth generation PCIe 4.0 data transmission interface that came with the next generation cards, which will certainly cause the reviewers of the hardware departments of the technical sites to confuse and certainly confuse the opinions of the users themselves regarding those who will buy new equipment and hardware during the coming period. While the fastest central processors currently used to measure the performance of new graphics cards are often the tenth-generation processors, such as the Core I9-10900K processor, but this central processor contains a PCIe 3.0 controller, and thus it may cause throttling of new graphics cards, even if the bottleneck percentage is Unremarkable and just very few tires, but the same logic does not apply to hardware reviewers.
The question now is, is it possible that conditions will change in the coming period and AMD processors begin for the first time to acquire the next graphics card reviews environment thanks to the RYZEN 3000 processors containing the built-in PCIe 4.0 controller, it seems to us that the matter is very logical and possible and will happen, which is what AMD hopes. There is no doubt that the only motherboards that support the PCIe 4.0 interface are the ones that contain the main chipset X570 and B550, unfortunately motherboards that contain Intel core chips that support the PCIe 4.0 interface cannot really take advantage of the new bus interface because the integrated controller for central processors of the PCIe generation 3.0 Only, so technical reviews will not rely on it unless the review is sponsored by Intel.
Let us also remember the advantages of the Core I9-10900K processor as we remember its disadvantages, although it does not have a PCIe 4.0 controller, but it is faster in performance compared to the R9 3950X processor by a number of points, especially with FHD 1080p display resolution thanks to its faster frequency. This means in the event that reviewers rely on the R9 3950X processor that there is a percentage of sacrifice in the level of performance with this display resolution (even if it is insignificant), but is it possible that the PCIe 4.0 interface is able to compensate for this slight difference, this is what we will see in Coming days.
If you want our personal opinion then let us tell you that it is better to stand on solid ground, in other words if you are planning to buy RTX 30 cards, you should also have a typical business platform that contains a PCIe 4.0 interface, and you will not find this support either by the central processor or the central chip except With AMD RYZEN platforms.
If you are a fan of the blue team then you should not worry, so far we are not sure what the results will look like and what can be achieved with the new PCIe 4.0 data transmission interface, but the new data transmission interface works with a bandwidth of 64GBps while the PCIe 3.0 transmission interface is limited. On 32GBps only. I think the crucial outcome is to wait to see results on the ground. Especially since the factory values of the data transfer interfaces are sometimes useless and useless with graphics cards in particular, unlike the NVMe PCIe SSD that benefits from any increase in bandwidth. So if you remember well, there was no noticeable difference between PCIe 2.0 and PCIe 3.0 when we first saw it on the Keplar GTX 600`s cards. At the time, the results did not differ whether these graphics cards were used with modern platforms that support the PCIe 3.0 interface or old platforms PCIe 2.0 or PCIe 1.1.
The idea and what is in it is that the larger bandwidth always allows for huge amounts of data to pass in the estimated time periods, which are calculated per second, but in most cases the graphics card is unable to process and produce this huge amount of data in the same period of time and therefore we do not notice a delay Or a difference, or in other words, a “bottleneck” between PCIe 3.0 vs PCIe 2.0 data transfer interface. But it must also be noted that the cards of the old generations were not as strong as the cards of the modern generations, so if the new bandwidth will make a difference in performance, this question is the answer to which lies in the near future.
New GDDR6X video memory
Both the RTX 3090 and RTX 3080 cards will come with GDDR6X video memory, and Nvidia’s promises have become clear that the new video memory developed by Micron factories will double the performance of GDDR6 memory, and thanks to the huge 24GB video memory size with the RTX 3090 processor and the size of 10GB with the RTX 3080 processor, they will have the advantage in capacity Download game data at better latency speeds.
It seems that Nvidia has bet on everything it has in the next generation cards destined for office devices, we will need to make sure that the central processors are able to keep up with these cards and not cause them to any percentage of the bottleneck, if you are seeking to own one of the RTX 30 cards higher category You should go to the Ryzen 9 or Ryzen 7 processors, but if you have an Intel platform, it is best to make sure you have a Core I7 or Core I9 processor. If you are planning to buy mid-range cards such as the RTX 3070, one of the mid-range processors such as the Ryzen 5 or Core I5 can be satisfied. Also, make sure that your case box is able to support the huge size of the RTX 3090 card, especially as it imposes three PCIe slots, and this means that this card may not fit with the mini-tower boxes, in addition to that, make sure to have a power supply of at least 750W. .
The new graphics cards will have many technical specifications and pioneering features in the field of graphics, and undoubtedly will bring with it a game experience full of excitement and fun and very close to real life, these graphics cards were and must come to meet the modern challenge imposed by the games of the current generation. It is worth noting that the new graphics cards will contain a bundled package of games that support modern graphic technologies. So far, we do not have much about the upcoming Ampere architecture cards, we will be patient with you so that we can get the answer that clarifies to us an explanation for all our questions, and certainly we will keep you updated with us on everything that is new and useful.