Main Page / Graphic cards / Comparision AMD Radeon RX 6900 XT vs NVIDIA GeForce RTX 4090
NVIDIA GeForce RTX 4090 NVIDIA GeForce RTX 4090
AMD Radeon RX 6900 XT AMD Radeon RX 6900 XT
VS

Comparision NVIDIA GeForce RTX 4090 vs AMD Radeon RX 6900 XT

NVIDIA GeForce RTX 4090

NVIDIA GeForce RTX 4090

Rating: 0 points
AMD Radeon RX 6900 XT

WINNER
AMD Radeon RX 6900 XT

Rating: 86 points
Grade
NVIDIA GeForce RTX 4090
AMD Radeon RX 6900 XT
Performance
8
8
Memory
2
8
General information
8
8
Functions
8
7
Ports
7
10

Top specs and features

GPU base clock speed

NVIDIA GeForce RTX 4090: 1815 MHz AMD Radeon RX 6900 XT: 1825 MHz

RAM

NVIDIA GeForce RTX 4090: 24 GB AMD Radeon RX 6900 XT: 16 GB

Memory bandwidth

NVIDIA GeForce RTX 4090: 1.018 GB/s AMD Radeon RX 6900 XT: 512 GB/s

Gpu memory speed

NVIDIA GeForce RTX 4090: 1325 MHz AMD Radeon RX 6900 XT: 2000 MHz

FLOPS

NVIDIA GeForce RTX 4090: 70.24 TFLOPS AMD Radeon RX 6900 XT: 23.75 TFLOPS

Description

The NVIDIA GeForce RTX 4090 video card is based on the Lovelace architecture. AMD Radeon RX 6900 XT on the RDNA 2.0 architecture. The first has There is no data million transistors. The second is 26800 million. NVIDIA GeForce RTX 4090 has a transistor size of 5 nm versus 7.

The base clock speed of the first video card is 1815 MHz versus 1825 MHz for the second.

Let's move on to memory. NVIDIA GeForce RTX 4090 has 24 GB. AMD Radeon RX 6900 XT has 24 GB installed. The bandwidth of the first video card is 1.018 Gb/s versus 512 Gb/s of the second.

FLOPS of NVIDIA GeForce RTX 4090 is 70.24. At AMD Radeon RX 6900 XT 23.75.

Goes to tests in benchmarks. In the Passmark benchmark, NVIDIA GeForce RTX 4090 scored There is no data points. And here is the second card 25762 points. In 3DMark, the first model scored There is no data points. Second 49079 points.

In terms of interfaces. The first video card is connected using There is no data. The second is PCIe 4.0 x16. Video card NVIDIA GeForce RTX 4090 has Directx version 12.2. Video card AMD Radeon RX 6900 XT -- Directx version - 12.2.

Regarding cooling, NVIDIA GeForce RTX 4090 has 450W heat dissipation requirements versus 300W for AMD Radeon RX 6900 XT.

Why AMD Radeon RX 6900 XT is better than NVIDIA GeForce RTX 4090

  • RAM 24 GB против 16 GB, more on 50%
  • FLOPS 70.24 TFLOPS против 23.75 TFLOPS, more on 196%
  • Technological process 5 nm против 7 nm, less by -29%

NVIDIA GeForce RTX 4090 vs AMD Radeon RX 6900 XT: highlights

NVIDIA GeForce RTX 4090
NVIDIA GeForce RTX 4090
AMD Radeon RX 6900 XT
AMD Radeon RX 6900 XT
Performance
GPU base clock speed
The graphics processing unit (GPU) has a high clock speed.
1815 MHz
max 2457
Average: 1124.9 MHz
1825 MHz
max 2457
Average: 1124.9 MHz
Gpu memory speed
This is an important aspect for calculating memory bandwidth.
1325 MHz
max 16000
Average: 1468 MHz
2000 MHz
max 16000
Average: 1468 MHz
FLOPS
Measuring the processing power of a processor is called FLOPS.
70.24 TFLOPS
max 1142.32
Average: 53 TFLOPS
23.75 TFLOPS
max 1142.32
Average: 53 TFLOPS
RAM
RAM in video cards (also known as video memory or VRAM) is a special type of memory used by a video card to store graphics data. It serves as a temporary buffer for textures, shaders, geometry, and other graphics resources that are needed to display images on the screen. More RAM allows the graphics card to work with more data and handle more complex graphic scenes with high resolution and detail. Show more
24 GB
max 128
Average: 4.6 GB
16 GB
max 128
Average: 4.6 GB
Number of threads
The more threads a video card has, the more processing power it can provide.
17408
max 18432
Average: 1326.3
max 18432
Average: 1326.3
Number of PCIe lanes
The number of PCIe lanes in video cards determines the speed and bandwidth of data transfer between the video card and other computer components through the PCIe interface. The more PCIe lanes a video card has, the more bandwidth and ability to communicate with other computer components. Show more
16
max 16
Average:
16
max 16
Average:
Pixel rendering speed
The higher the pixel rendering speed, the smoother and more realistic the display of graphics and the movement of objects on the screen will be.
370 GTexel/s    
max 563
Average: 94.3 GTexel/s    
288 GTexel/s    
max 563
Average: 94.3 GTexel/s    
TMUs
Responsible for texturing objects in 3D graphics. TMU provides textures to the surfaces of objects, which gives them a realistic look and detail. The number of TMUs in a video card determines its ability to process textures. The more TMUs, the more textures can be processed at the same time, which contributes to better texturing of objects and increases the realism of graphics. Show more
544
max 880
Average: 140.1
320
max 880
Average: 140.1
ROPs
Responsible for the final processing of pixels and their display on the screen. ROPs perform various operations on pixels, such as blending colors, applying transparency, and writing to the framebuffer. The number of ROPs in a video card affects its ability to process and display graphics. The more ROPs, the more pixels and image fragments can be processed and displayed on the screen at the same time. A higher number of ROPs generally results in faster and more efficient graphics rendering and better performance in games and graphics applications. Show more
192
max 256
Average: 56.8
128
max 256
Average: 56.8
Number of shader blocks
The number of shader units in video cards refers to the number of parallel processors that perform computational operations in the GPU. The more shader units in the video card, the more computing resources are available for processing graphics tasks. Show more
17408
max 17408
Average:
5120
max 17408
Average:
L2 cache size
Used to temporarily store data and instructions used by the graphics card when performing graphics calculations. A larger L2 cache allows the graphics card to store more data and instructions, which helps speed up the processing of graphics operations. Show more
96000
4000
Turbo gpu
If the GPU speed has dropped below its limit, then to improve performance, it can go to a high clock speed.
1925 MHz
max 2903
Average: 1514 MHz
2250 MHz
max 2903
Average: 1514 MHz
architecture name
Lovelace
RDNA 2.0
GPU name
AD102
Navi 21
Memory
Memory bandwidth
This is the rate at which the device stores or reads information.
1.018 GB/s
max 2656
Average: 257.8 GB/s
512 GB/s
max 2656
Average: 257.8 GB/s
RAM
RAM in video cards (also known as video memory or VRAM) is a special type of memory used by a video card to store graphics data. It serves as a temporary buffer for textures, shaders, geometry, and other graphics resources that are needed to display images on the screen. More RAM allows the graphics card to work with more data and handle more complex graphic scenes with high resolution and detail. Show more
24 GB
max 128
Average: 4.6 GB
16 GB
max 128
Average: 4.6 GB
GDDR memory versions
Latest versions of GDDR memory provide high data transfer rates to improve overall performance
6
max 6
Average: 4.9
6
max 6
Average: 4.9
Memory bus width
A wide memory bus means that it can transfer more information in one cycle. This property affects memory performance as well as the overall performance of the device's graphics card. Show more
384 bit
max 8192
Average: 283.9 bit
256 bit
max 8192
Average: 283.9 bit
General information
Crystal size
The physical dimensions of the chip on which the transistors, microcircuits and other components necessary for the operation of the video card are located. The larger the die size, the more space the GPU takes up on the graphics card. Larger die sizes can provide more computing resources such as CUDA cores or tensor cores, which can result in increased performance and graphics processing capabilities. Show more
611
max 826
Average: 356.7
520
max 826
Average: 356.7
Length
334
max 524
Average: 250.2
267
max 524
Average: 250.2
Generation
A new generation of graphics card usually includes improved architecture, higher performance, more efficient use of power, improved graphics capabilities, and new features. Show more
GeForce 40
Navi II
Manufacturer
TSMC
TSMC
Power supply power
When choosing a power supply for a video card, you must take into account the power requirements of the video card manufacturer, as well as other computer components. Show more
850
max 1300
Average:
700
max 1300
Average:
Year of issue
2022
max 2023
Average:
2020
max 2023
Average:
Power Consumption (TDP)
Heat Dissipation Requirements (TDP) is the maximum possible amount of energy dissipated by the cooling system. The lower the TDP, the less power will be consumed Show more
450 W
Average: 160 W
300 W
Average: 160 W
Technological process
The small size of the semiconductors means this is a new generation chip.
5 nm
Average: 34.7 nm
7 nm
Average: 34.7 nm
PCIe connection interface
A considerable speed of the expansion card used to connect the computer to the peripherals is provided. The updated versions offer impressive bandwidth and high performance. Show more
4
max 4
Average: 3
4
max 4
Average: 3
Width
142 mm
max 421.7
Average: 192.1 mm
120 mm
max 421.7
Average: 192.1 mm
Height
63 mm
max 620
Average: 89.6 mm
52 mm
max 620
Average: 89.6 mm
Purpose
Desktop
Desktop
Functions
OpenGL Version
OpenGL provides access to the graphics card's hardware capabilities for displaying 2D and 3D graphics objects. New versions of OpenGL may include support for new graphical effects, performance optimizations, bug fixes, and other improvements. Show more
4.6
max 4.6
Average:
4.6
max 4.6
Average:
DirectX
Used in demanding games, providing improved graphics
12.2
max 12.2
Average: 11.4
12.2
max 12.2
Average: 11.4
Shader model version
The higher the version of the shader model in the video card, the more functions and possibilities are available for programming graphic effects.
6.6
max 6.7
Average: 5.9
6.5
max 6.7
Average: 5.9
CUDA Version
Allows you to use the compute cores of your graphics card to perform parallel computing, which can be useful in areas such as scientific research, deep learning, image processing, and other computationally intensive tasks. Show more
9
max 9
Average:
max 9
Average:
Ports
Number of connectors 16-pin
1
max 1
Average: 1
max 1
Average: 1
Has hdmi output
HDMI output allows you to connect devices with HDMI or mini HDMI ports. They can send video and audio to the display.
Available
Available
HDMI version
The latest version provides a wide signal transmission channel due to the increased number of audio channels, frames per second, etc.
2.1
max 2.1
Average: 1.9
2.1
max 2.1
Average: 1.9
Number of HDMI connectors
The more their number, the more devices can be connected at the same time (for example, game / TV set-top boxes)
1
max 3
Average: 1.1
1
max 3
Average: 1.1
HDMI
A digital interface that is used to transmit high-resolution audio and video signals.
Available
Available

FAQ

How does the NVIDIA GeForce RTX 4090 processor perform in benchmarks?

Passmark NVIDIA GeForce RTX 4090 scored There is no data points. The second video card scored 25762 points in Passmark.

What FLOPS do video cards have?

FLOPS NVIDIA GeForce RTX 4090 is 70.24 TFLOPS. But the second video card has FLOPS equal to 23.75 TFLOPS.

What power consumption?

NVIDIA GeForce RTX 4090 450 Watt. AMD Radeon RX 6900 XT 300 Watt.

How fast are NVIDIA GeForce RTX 4090 and AMD Radeon RX 6900 XT?

NVIDIA GeForce RTX 4090 operates at 1815 MHz. In this case, the maximum frequency reaches 1925 MHz. The clock base frequency of AMD Radeon RX 6900 XT reaches 1825 MHz. In turbo mode it reaches 2250 MHz.

What kind of memory do graphics cards have?

NVIDIA GeForce RTX 4090 supports GDDR6. Installed 24 GB of RAM. Throughput reaches 1.018 GB/s. AMD Radeon RX 6900 XT works with GDDR6. The second one has 16 GB of RAM installed. Its bandwidth is 1.018 GB/s.

How many HDMI connectors do they have?

NVIDIA GeForce RTX 4090 has 1 HDMI outputs. AMD Radeon RX 6900 XT is equipped with 1 HDMI outputs.

What power connectors are used?

NVIDIA GeForce RTX 4090 uses There is no data. AMD Radeon RX 6900 XT is equipped with There is no data HDMI outputs.

What architecture are video cards based on?

NVIDIA GeForce RTX 4090 is built on Lovelace. AMD Radeon RX 6900 XT uses the RDNA 2.0 architecture.

What graphics processor is being used?

NVIDIA GeForce RTX 4090 is equipped with AD102. AMD Radeon RX 6900 XT is set to Navi 21.

How many PCIe lanes

The first graphics card has 16 PCIe lanes. And the PCIe version is 4. AMD Radeon RX 6900 XT 16 PCIe lanes. PCIe version 4.

How many transistors?

NVIDIA GeForce RTX 4090 has There is no data million transistors. AMD Radeon RX 6900 XT has 26800 million transistors

Graphic cards