Home > Graphics cards > The best 10 graphics cards in comparison

The best 10 graphics cards in comparison

Graphics cards (1 - 5)
Asus ROG Matrix GeForce RTX 4090 PlatinumAsus TUF GeForce RTX 4090Colorful iGame GeForce RTX 4090 Neptune OCMSI GeForce RTX 4090 Suprim Liquid XMSI GeForce RTX 4090 Suprim LiquidAsus ROG Strix GeForce RTX 4090Inno3D iChill GeForce RTX 4090 FrostbiteAsus ROG Strix GeForce RTX 4090 White OC EditionAsus TUF GeForce RTX 4090 OCAsus ROG Strix GeForce RTX 4090 White Edition
Image
Asus ROG Matrix GeForce RTX 4090 Platinum
Asus TUF GeForce RTX 4090
Colorful iGame GeForce RTX 4090 Neptune OC
MSI GeForce RTX 4090 Suprim Liquid X
MSI GeForce RTX 4090 Suprim Liquid
Asus ROG Strix GeForce RTX 4090
Inno3D iChill GeForce RTX 4090 Frostbite
Asus ROG Strix GeForce RTX 4090 White OC Edition
Asus TUF GeForce RTX 4090 OC
Asus ROG Strix GeForce RTX 4090 White Edition
Best price
Best price
Performance
GPU clock speedThe clock speed of the graphics processing unit (GPU).
GPU clock speedThe clock speed of the graphics processing unit (GPU).2235 MHz
2235 MHz
2235 MHz
2235 MHz
2235 MHz
2235 MHz
2235 MHz
2235 MHz
2235 MHz
2235 MHz
Pixel rateThe number of pixels that can be rendered to the screen every second.
Pixel rateThe number of pixels that can be rendered to the screen every second.469.9 GPixel/s
443.5 GPixel/s
464.6 GPixel/s
462 GPixel/s
456.7 GPixel/s
443.5 GPixel/s
454.1 GPixel/s
459.4 GPixel/s
451.44 GPixel/s
443.5 GPixel/s
Floating-point performanceFloating-point performance is a measurement of the raw processing power of the GPU.
Floating-point performanceFloating-point performance is a measurement of the raw processing power of the GPU.87.49 TFLOPS
82.58 TFLOPS
86.51 TFLOPS
86.02 TFLOPS
85.03 TFLOPS
82.58 TFLOPS
84.54 TFLOPS
85.52 TFLOPS
84.05 TFLOPS
82.58 TFLOPS
Texture rateThe number of textured pixels that can be rendered to the screen every second.
Texture rateThe number of textured pixels that can be rendered to the screen every second.1367 GTexels/s
1290 GTexels/s
1352 GTexels/s
1344 GTexels/s
1329 GTexels/s
1290 GTexels/s
1321 GTexels/s
1336 GTexels/s
1313.28 GTexels/s
1290 GTexels/s
GPU memory speedThe memory clock speed is one aspect that determines the memory bandwidth.
GPU memory speedThe memory clock speed is one aspect that determines the memory bandwidth.1313 MHz
1325 MHz
1325 MHz
1325 MHz
1325 MHz
1325 MHz
1313 MHz
1313 MHz
1325 MHz
1313 MHz
Total score for "Performance"
Total score for "Performance"
Memory
Effective memory speedThe effective memory clock speed is calculated from the size and data rate of the memory. Higher clock speeds can give increased performance in games and other apps.
Effective memory speedThe effective memory clock speed is calculated from the size and data rate of the memory. Higher clock speeds can give increased performance in games and other apps.21000 MHz
21000 MHz
21000 MHz
21000 MHz
21000 MHz
21000 MHz
21000 MHz
21000 MHz
21000 MHz
21000 MHz
Maximum memory bandwidthThis is the maximum rate that data can be read from or stored into memory.
Maximum memory bandwidthThis is the maximum rate that data can be read from or stored into memory.1008 GB/s
1018 GB/s
1008 GB/s
1018 GB/s
1018 GB/s
1018 GB/s
1008 GB/s
1008 GB/s
1018 GB/s
1008 GB/s
VRAMVRAM (video RAM) is the dedicated memory of a graphics card. More VRAM generally allows you to run games at higher settings, especially for things like texture resolution.
VRAMVRAM (video RAM) is the dedicated memory of a graphics card. More VRAM generally allows you to run games at higher settings, especially for things like texture resolution.24GB
24GB
24GB
24GB
24GB
24GB
24GB
24GB
24GB
24GB
GDDR versionNewer versions of GDDR memory offer improvements such as higher transfer rates that give increased performance.
GDDR versionNewer versions of GDDR memory offer improvements such as higher transfer rates that give increased performance.GDDR6X
GDDR6X
GDDR6X
GDDR6X
GDDR6X
GDDR6X
GDDR6X
GDDR6X
GDDR6X
GDDR6X
Memory bus widthA wider bus width means that it can carry more data per cycle. It is an important factor of memory performance, and therefore the general performance of the graphics card.
Memory bus widthA wider bus width means that it can carry more data per cycle. It is an important factor of memory performance, and therefore the general performance of the graphics card.384-bit
384-bit
384-bit
384-bit
384-bit
384-bit
384-bit
384-bit
384-bit
384-bit
Total score for "Memory"
Total score for "Memory"
Features
OpenGL versionOpenGL is used in games, with newer versions supporting better graphics.
OpenGL versionOpenGL is used in games, with newer versions supporting better graphics.4.6
4.6
4.6
4.6
4.6
4.6
4.6
4.6
4.6
4.6
OpenCL versionSome apps use OpenCL to apply the power of the graphics processing unit (GPU) for non-graphical computing. Newer versions introduce more functionality and better performance.
OpenCL versionSome apps use OpenCL to apply the power of the graphics processing unit (GPU) for non-graphical computing. Newer versions introduce more functionality and better performance.3
3
3
3
3
3
3
3
3
3
Multi-display technologyThe graphics card supports multi-display technology. This allows you to configure multiple monitors in order to create a more immersive gaming experience, such as having a wider field of view.
Multi-display technologyThe graphics card supports multi-display technology. This allows you to configure multiple monitors in order to create a more immersive gaming experience, such as having a wider field of view.
Ray tracingRay tracing is an advanced light rendering technique that provides more realistic lighting, shadows, and reflections in games.
Ray tracingRay tracing is an advanced light rendering technique that provides more realistic lighting, shadows, and reflections in games.
DLSSDLSS (Deep Learning Super Sampling) is an upscaling technology powered by AI. It allows the graphics card to render games at a lower resolution and upscale them to a higher resolution with near-native visual quality and increased performance. DLSS is only available on select games.
DLSSDLSS (Deep Learning Super Sampling) is an upscaling technology powered by AI. It allows the graphics card to render games at a lower resolution and upscale them to a higher resolution with near-native visual quality and increased performance. DLSS is only available on select games.
Total score for "Features"
Total score for "Features"
Ports
HDMI outputDevices with a HDMI or mini HDMI port can transfer high definition video and audio to a display.
HDMI outputDevices with a HDMI or mini HDMI port can transfer high definition video and audio to a display.
HDMI portsMore HDMI ports mean that you can simultaneously connect numerous devices, such as video game consoles and set-top boxes.
HDMI portsMore HDMI ports mean that you can simultaneously connect numerous devices, such as video game consoles and set-top boxes.2
2
1
1
1
2
1
2
2
2
HDMI versionNewer versions of HDMI support higher bandwidth, which allows for higher resolutions and frame rates.
HDMI versionNewer versions of HDMI support higher bandwidth, which allows for higher resolutions and frame rates.HDMI 2.1
HDMI 2.1a
HDMI 2.1
HDMI 2.1
HDMI 2.1
HDMI 2.1
HDMI 2.1
HDMI 2.1a
HDMI 2.1a
HDMI 2.1a
DisplayPort outputsAllows you to connect to a display using DisplayPort.
DisplayPort outputsAllows you to connect to a display using DisplayPort.3
3
3
3
3
3
3
3
3
3
Total score for "Ports"
Total score for "Ports"
General info
Thermal Design Power (TDP)The thermal design power (TDP) is the maximum amount of power the cooling system needs to dissipate. A lower TDP typically means that it consumes less power.
Thermal Design Power (TDP)The thermal design power (TDP) is the maximum amount of power the cooling system needs to dissipate. A lower TDP typically means that it consumes less power.450W
450W
450W
450W
450W
450W
450W
450W
450W
450W
Number of transistorsA higher transistor count generally indicates a newer, more powerful processor.
Number of transistorsA higher transistor count generally indicates a newer, more powerful processor.76300 million
76300 million
76300 million
76300 million
76300 million
76300 million
76300 million
76300 million
76300 million
76300 million
Semiconductor sizeSmall semiconductors provide better performance and reduced power consumption. Chipsets with a higher number of transistors, semiconductor components of electronic devices, offer more computational power. A small form factor allows more transistors to fit on a chip, therefore increasing its performance.
Semiconductor sizeSmall semiconductors provide better performance and reduced power consumption. Chipsets with a higher number of transistors, semiconductor components of electronic devices, offer more computational power. A small form factor allows more transistors to fit on a chip, therefore increasing its performance.5 nm
4 nm
4 nm
4 nm
4 nm
4 nm
4 nm
4 nm
4 nm
4 nm
PCI Express (PCIe) versionPeripheral Component Interconnect Express (PCIe) is a high-speed interface standard for connecting components, such as graphics cards and SSDs, to a motherboard. Newer versions can support more bandwidth and deliver better performance.
PCI Express (PCIe) versionPeripheral Component Interconnect Express (PCIe) is a high-speed interface standard for connecting components, such as graphics cards and SSDs, to a motherboard. Newer versions can support more bandwidth and deliver better performance.4
4
4
4
4
4
4
4
4
4
Warranty periodWhen covered under the manufacturer’s warranty it is possible to get a replacement in the case of a malfunction. Note: This may vary by region.
Warranty periodWhen covered under the manufacturer’s warranty it is possible to get a replacement in the case of a malfunction. Note: This may vary by region.N.A.3 years
N.A.N.A.N.A.N.A.N.A.N.A.N.A.N.A.
Total score for "General info"
Total score for "General info"

How to choose the best graphics card for your computer

For gaming and professional work with graphics software, such as animation and video editing, the graphics card is one of the essential components of a computer. For such workloads, the integrated graphics of most new CPUs won't be enough, and you'll have to add a graphics card to your system. With a wide range of GPUs to choose from, it's crucial to know which specifications and features are the most important. To help you find the best graphics card for your needs, we compiled a list of tips.

Defining your goal and budget

Before buying a graphics card, you first have to determine what kind of workloads you are primarily dealing with and how much you are willing to spend. What applications are you typically running on your computer? Are you mostly browsing the internet, or are you often playing games with amazing graphics, like Fallout or Battlefied? Are you editing videos and photos or watching movies in 4K high-definition resolution?

There are two main types of graphics cards: gaming-oriented and workstation-oriented. The gaming-oriented GPUs usually support very high resolutions and frame rates. In contrast, the workstation-oriented GPUs (meant for using advanced photo and video editing software or 3D modeling software, for example) have high texture and pixel rates.

While graphics cards have a wide price range, each tier has a specific price range for its group. The low-end graphics cards usually cost less than $300, while the mid-tier cards tend to stick around $400-$500, and the high-end cards go well beyond $800.

Key specs and features of GPUs

Memory Type (VRAM)

The memory type of the graphics card is directly related to the performance of the card. Though clock speeds matter, the memory or VRAM matters a bit more. Take, for example, graphics cards with GDDR5x versus graphics cards with GDDR5. The GDDR5x memory has 64 bits/access data rate compared to 32 bits/access on the GDDR5. In 2018, the first cards with GDDR6, with transfer rates of up to 128 GB/s, entered the market. Some cards use a different memory type called High Bandwidth Memory (HBM), which is smaller than that of the GDDR5x but delivers quite the boost in performance at 1024 bits/access.

One of the reasons why manufacturers have opted out of HBM, for now, is that this technology uses an entirely different way of processing and manufacturing. GDDR5 and GDDR5x have the same physical sizes, thus manufacturing time and cost almost remain the same. HBM has a smaller form factor, and this means manufacturers have to change their process to cater to this technology.

Memory type Speed Bandwidth
GDDR5 32-bit/access 8 Gbit/s per pin
GDDR5x 64-bit/access 14 Gbit/s per pin
GDDR6 64-bit/access 16 Gbit/s per pin
HBM 1024-bit/access 128 Gbit/s per pin

GPU clock speed

The clock speed determines the performance of a graphics card, as explained in detail here. There are two types: the base clock and the boost clock. The base clock is the actual speed at which a card can operate, while the boost clock is the highest guaranteed performance. Some users tend to overclock their cards beyond the rated clock speeds to achieve a faster and better performance for graphics-intensive applications.

Shading Units (steam processors or CUDA)

Shading units, shader units, steam processors, or CUDA cores, are small processors on the cards that process data, and a card can contain thousands of these. These shading units help render images, especially when three-dimensional scenes are at play. Games and video editing software benefit from a high number of shading units because they help process the thousands of frames the software has to render.

Resolution

Most cards to date can provide high definition resolution at 1080p, while more expensive graphics cards, such as the RTX 2080 series, enable you to enjoy 4K resolution. To fully enjoy 4K resolution, you also need a monitor with high resolution. You can read more about display resolution here.

Cooling

Graphics cards tend to use quite a lot of power, and this translates into heat. As with any electronic device, heat is the enemy of graphics cards, and it may just take a few seconds to a minute for a card to fry without proper cooling. Though the integrated stock coolers usually provide adequate cooling, some cards are equipped with additional coolers, for example, two or three large fans that help remove heat. Other manufacturers even place specialized cooling solutions that use water to cool the GPU. Each cooling solution delivers different results and requires specific maintenance. The added features cost more than the stock cooling solution that is placed in all reference cards.

AMD or Nvidia

This is a tough question to answer, just like Windows versus Mac, Android versus iOS, and other challenging comparisons. What is interesting with these two companies is their development and process to support virtual reality (VR). Recently, AMD graphics cards have been released that aim to bring the VR experience at a low cost compared to its competition. Nvidia, however, brings quite a different appeal to the VR market through high performance throughout their entire GPU range. Nvidia graphics cards cater not just to VR but also to other applications, which explains the higher price tag.

By Don Chen
| Updated on
This page is currently only available in English.