NVIDIA
Finally, the GEFORCE RTX 4080 graphics card has reached us from Nvidia, which is a new high -end category card and the second most powerful card in the new generation of Nvidia Graphic Cards in ADA LOVEACE architecture with the accuracy of 4 nm of Tsmc Taiwan, which is an improved manufacturing accuracy of the accuracy of 5 nm. This card is built on the AD103 chip and not the AD102 chip that the RTX upper category card uses 4090. The size of this chip reaches about two -thirds of the AD102 size and it includes about 45.9 billion transistors. Nvidia cards bearing the number XX80 are cards that have a great legacy in the world of graphic cards and we do not doubt that many are interested in the results of examination and inspection of this card and whether it maintains the ancient history of this series in advanced performance and breaking expectations.
Learn about the new generation RTX 4080
The 16GB card of GDDR6X RAM with 256-bit data interface. Two years ago, when Nvidia launched the RTX30 cards with the Aprice architecture, these memory units were only available with a capacity of 1GB for each of the manufacturing company and the maximum that could be used on the card panel 12 pieces or unit (RTX 3090 Card contained 12 and a unit on each side of the electronic plate). After that, the units are available with a capacity of 2GB per unit, and Nvidia can provide 16GB of memory with only 8 units, and they operate at 22.4 GBPS, and the total flow of memory reaches 716 GB per second.
These cards do not use all the processors on the chip, as they are satisfied with 76 shading units out of 80 units, as Nvidia has disabled 4 shading units, which means that the card uses 9728 CUDA nucleus instead This chip when the production process improves. The card operates at a basic frequency of 2.2 GHz and 2.5 GHz with payment technology and these numbers are much higher than the last generation in Aprier One of the fastest graphics cards in the frequency throughout history. Finally, the card includes 304 Tensor units to process artificial intelligence and 76 processors to accelerate radiology tracking accounts. The power of the card is 49 teraflops.
RTX 3080 vs RTX 4080
The card consumes 320 watts of electric power and this is identical to the last generation RTX 3080 and less than the RTX 4090 that consumes 450 watts of electric power, but the card uses a new energy port with 12vHPWR connection that supports the new standard specifications of ATX 3.0 which has a number of Additional advantages compared to the previous, such as the ability to support graphic cards with an electric card, up to 600 watts, in addition to avoiding the problems of electrical energy distribution to a number of previous 8-PIN connections. There are two companies that make new connections, the two companies from Taiwan and NTK (used by companies such as Zotac and Gigabyte) and ASRON used by NVIDIA itself.
The connection that comes with this card in the cartoon from Nvidia includes a conversion to 3 8-PIN ports so that the owners of old energy supplies can connect these cards and provide them with the necessary electric energy, the real problem with regard to the connection is in the location of the electric power port in the card and it comes on the side, The distance was not sufficient in the box to connect the port with the transfer, as well as 8-PIN wires from the side, and therefore we had to keep the computer box open. You should make sure that you have the depth of the box if you wish to acquire this card.
This card maintains the usual outlets of Nvidia that were present in the last generation cards, with HDMI 2.1 and three DP 1.4 ports, and the card does not offer BIOS options as the third-party cards did from Ampre in the past generation, but it does not really need it because it offers a mixture Ideal of electric power consumption, noise and performance and we will come to mention this later.
With regard to the price, the official price of this card in the West is $ 1,200, and it exceeds the stages of the RTX 3080 card that started at $ 700, but it is a fact that this price only has for small periods. One of the direct reasons for the high price is the increase in production costs of manufacturing accuracy, as the cost of one disk (WAFer) reaches more than $ 17,000, which is much more expensive than in the past, and this is also confirmed to us by the president of Nvidia earlier. Of course, Nvidia also undoubtedly noticed the great demand for graphic cards in the past generation at high prices, and this is also also playing his role in the new pricing. However, the card comes very thick and significantly greater than the past generation (about 3 spaces or Slots).
