GeForce RTX 3080 and Radeon RX 6800 XT are two premium Nvidia and AMD video cards that promise high performance when playing. Launched in 2020, the GPUs bring the new Ampère and RDNA2 architectures, respectively, delivering features such as native Ray Tracing and, in the case of the Nvidia option, DLSS 2.0. Also, manufacturers speak of sufficient capacity to run games at 4K at 60 fps, surpassing the top-of-the-line versions of previous generations.
Compare below the technical sheet of both models and find out which powerful video card is more worthwhile for you.
GeForce RTX 3080 vs Radeon RX 6800 XT
|Specifications||GeForce RTX 3080||Radeon RX 6800 XT|
|Launch||September 2020||October 2020|
|Cores||8,704 CUDA cores||4,608 stream processors|
|Velocity||1,440 to 1,710 MHz||1,825 to 2,250 MHz|
|RAM||16 GB GDDR6||16 GB GDDR6|
|Memory speed||19 Gb / s||16 Gb / s|
|Memory interface||320 bits||256 bits|
|Bandwidth||760 GB / s||512 GB / s|
|TBP||320 Watts||300 Watts|
One of AMD’s high stakes in the premium graphics card market, the Radeon RX 6800 XT is a direct competitor to Nvidia’s RTX 3080. The GPU consists of a 72-unit graphics processor with 4,608 stream processors. The Nvidia model is formed by 8,704 CUDA processing cores.
At speeds, Radeon has clocks of 1,825 MHz for standard speed, 2,015 MHz clock for games (the speed that AMD advocates to be stable), and can reach 2,250 MHz with turbo activated. On the Nvidia side, the typical RTX 3080 should operate at 1,410 MHz standard speed and 1,710 MHz with turbo activated. Taking into account the datasheet numbers, the higher speeds of Radeon outweigh the fact that the AMD board has less processing cores.
Concerning memory, the RX 6800 XT adds 16 GB of GDDR6 that works at 16 Gb/s (gigabits per second) and, connected to the graphics processor via a 256-bit interface, it uses a 510 GB/s band. The Nvidia board has 10 GB of GDDR6X that works at 19 Gb/s, has a 320-bit interface, and a bandwidth of 760 GB/s.
Of all these numbers, the most important is the bandwidth. The Nvidia product exchanges 760 GB of data per second with the graphics processor, while the Radeon has 510 GB. However, AMD’s architecture is more efficient in the way it works with memory, so this difference ends up being mitigated.
Another important feature of both is the fact that both offer specialized hardware for accelerating Ray Tracing effects in real-time. In the case of AMD, there are 72 cores for RT, while the Nvidia model comes with 68.
AMD’s board is competitive in performance and features, but first comparisons and reviews indicate some advantages for Nvidia. The RTX 3080 is faster when it comes to running games in 4K and displays higher performance with Ray Tracing enabled. Also, it features DLSS 2.0, which uses image reconstruction by artificial intelligence to run games in high graphics fidelity with a lower performance cost. This is a technology that AMD is not yet competing with.
Radeon’s counterparts are also important: the RX 6800 XT has more RAM, something important for those who want to play in 4K.
The PC Gamer site has made some comparisons that show the RX 6800 XT getting very close to the rival when running games in 4K. With Metro Exodus, recognized for the excellent use of Ray Tracing, the board from AMD yields an average of 66 fps in 4K with graphics on Ultra, while the GeForce RTX 3080 goes a little further and beats 71 fps in the same conditions. The difference was almost the same in comparison using Horizon: Zero Dawn: 71 fps for Radeon and 77 fps for GeForce in gameplay with 4K resolution and maximum graphics.
In F1 2020, reaching 141 fps average running the game in 4K with graphics on Ultra against 135 fps of Nvidia’s rival. Comparisons of other sites reinforce the trend that the two boards are very close, either running in 4K or 1440p. However, the Nvidia board still has a little more margin due to higher performance, better performance with Ray Tracing, and the important differential of DLSS.
According to Nvidia, the RTX 3080 has a TBP (an acronym for “Total Board Power”) of 320 Watts, a number that reflects the amount of energy the product needs to operate in high demand. It is worth remembering that the name is different from the TDP value, associated with how much energy the board dissipates as heat during use.
For the Radeon board, AMD estimates the value of TBP at 300 Watts. In practice, this allows us to state that the GPU tends to be more economical, which explains why the RX 6800 XT tends to heat less.
The two brands also make their recommendations regarding power supply. For Nvidia, the ideal is that your system has, at least, a source of 750 Watts of real power, the same recommendation made by AMD.
Price and availability
While the GeForce RTX 3080 appears at prices for $1,374.99 in the national retail, the GPU of AMD still does not appear frequently in stores
The new Radeon has entered the order phase in the specialized stores. This is the same problem of high demand and shortage faced by Nvidia since the launch of GeForce RTX 30, as well as by Sony and Microsoft with the new generation of consoles.
Radeon versions have been offered for sale on the domestic market in the range of $1,649.
Partner brands of both manufacturers have already put their versions on the market. In terms of Radeon, there is not much to choose from: so far, the versions available are, Saphire, XFX RX, Asus, Gigabyte, and MSI, all arrive with the same standard AMD clock and without any overclock.
For Nvidia’s GPU, the maneuvering margins are higher. ZOTAC, ASUS, MSI, Gigabyte, and EVGA models can bring factory overclock, while there are more well-behaved versions of these same brands and smaller impact manufacturers.
This post may contain affiliate links, which means that I may receive a commission if you make a purchase using these links. As an Amazon Associate, I earn from qualifying purchases.