Picking the right graphics card for video editing needs can be daunting, especially with the constant innovations and releases from both Nvidia and AMD. We know that both companies have been at the forefront of graphics technology, pushing the limits of what’s possible and competing to offer the best solutions for various use cases.
As a video editor or content creator, your main concern is finding a GPU that meets your performance needs and fits within your budget.
So, Nvidia or AMD?
After doing some deep research, we came to the following conclusion:
Nvidia is generally better than AMD for video editing for several reasons: CUDA technology, software compatibility, and more optimized, consistent drivers. However, some AMD cards are better than Nvidia’s.
Ok, you have a short answer, but reading only that is not enough to make the best decision, so keep reading.
Here is why Nvidia is better overall
As I was going through my research by reading various reviews, articles, and recommendations of editors on forums, I noticed that 90% of them recommend Nvidia. To be honest, I was not expecting such a high percentage. However, things started to make sense when they presented their arguments.
Here are three that stick out the most:
CUDA technology
Nvidia’s GPUs utilize CUDA (Compute Unified Device Architecture) cores, which are specifically designed for parallel processing. CUDA technology can provide significant performance boosts in video editing applications that support GPU acceleration, such as Adobe Premiere Pro and DaVinci Resolve. In comparison, AMD GPUs use OpenCL, which is not as widely supported or optimized for video editing tasks.
OpenCL is the issue with AMD GPUs for video editing. OpenCL is now at version 3.0, although only Nvidia and Intel support it.
OpenCL support on the latest AMD GPUs is stuck at version 2.0 (2.1 for the RX 6000 series), and OpenCL 2.x has proven difficult for video editing software regarding GPGPU acceleration.
As a result, on an AMD GPU-powered system, some rendering features that would normally have gone to the GPU on an Nvidia GPU-powered system instead went straight to the CPU.
I noticed the artifacts in output on AMD encodes are far more common, which is probably why some people say, “They probably used an AMD card for this one” on some Youtube videos.
Software compatibility
Most popular video editing applications have been optimized to work seamlessly with Nvidia GPUs. For example, Adobe’s Mercury Playback Engine, which powers Adobe Premiere Pro, is designed to work well with Nvidia’s CUDA technology. While AMD GPUs also offer support for these applications, the optimization level may not be as high.
For instance, the challenge with DaVinci Resolve is that numerous workflows don’t support OpenCL, as they rely exclusively on CUDA. Consequently, if you’re utilizing anything beyond basic features, you’ll likely need an NVIDIA graphics card for optimal performance.
A similar situation exists with Adobe Premiere, and this doesn’t even take into account plugins, which predominantly use CUDA. Currently, there’s a noticeable lack of new OpenCL-based developments. Since AMD doesn’t offer an alternative to CUDA, it’s unlikely that the situation will improve significantly, particularly on Windows, where OpenCL support is limited.
As we saw in the Puget Systems review, Nvidia cards performed better than AMD Radeon cards in Premiere Pro.
AMD is encouraging users to transition to ROCm and HIP. However, the list of compatible GPUs is quite small, and these platforms aren’t available on Windows. This further highlights the challenges AMD faces in providing competitive alternatives to NVIDIA’s CUDA ecosystem.
From what I have heard, AMD GPUs seem to do very well in Vegas overall, but Nvidia is still better in terms of playback performance. If you check their official site, their Movie Studio states solid support for AMD cards.
To summarize, if your video editing program supports nVidia’s proprietary CUDA GPGPU system, the nVidia card is definitely a superior choice. If the program is more friendly toward the OpenCL system, AMD might a better choice.
Consistent driver updates
Nvidia is known for providing frequent driver updates, ensuring compatibility and optimization for various video editing applications. While AMD has improved its driver support recently, some users still report occasional compatibility issues and performance inconsistencies.
AMD was floundering and on the verge of bankruptcy in the early 2010s, but they recovered spectacularly by the end of the decade. As a result, AMD had very little money to invest, and they spent it largely on building better hardware.
Nvidia had a lot of money, so they invested in other features like RTX and improved drivers. Nvidia’s more frequent and optimized drivers give it a significant advantage for streaming, video editing, and various other professional tasks ranging from graphic design to 3D modeling.
Some editors say that they even like AMD cards, but due to their software support, they simply go with Nvidia. Nowadays, AMD is generally the greatest bargain option for pure gaming, though this will always change over time.
From what I can conclude, the CUDA is the major reason. As we can see, AMD simply lacks this. AMD has been the weakest GPU manufacturer for a decade, but they are now making cards that can compete with Nvidia. CUDA provides a more robust pre-built infrastructure than AMD’s core processors. Also, Nvidia is investing more in CUDA software development than AMD.
The only case I can think of for AMD is a higher VRAM amount in the 6000 series vs RTX 30XX. As we know, having more VRAM is often better for video editing.
Head-to-head comparison
To actually provide you with some proven data, we had to visit Puget Systems.
First started off with their newest Nvidia vs AMD comparison.
The GPUs tested in this review were:
NVIDIA GeForce RTX 4090 24GB
NVIDIA GeForce RTX 4080 16GB
Asus GeForce RTX 4070 Ti STRIX 12GB
NVIDIA GeForce RTX 3090 Ti 24GB
NVIDIA GeForce RTX 3090 24GB
NVIDIA GeForce RTX 3080 Ti 12GB
NVIDIA GeForce RTX 3080 10GB
AMD Radeon RX 7900 XTX 24GB
AMD Radeon RX 6900 XT 16GB
Intel Arc A770 16GB
CPU: AMD Threadripper Pro 5975WX 32 core
MOBO: Asus Pro WX WRX80E-SAGE
RAM: 128GB
PSU: 1600W
OS: Windows 11 Pro 64-bit
Here are some results from DaVinci Resolve:
Red/Braw Result Geoman
| GPU | Score |
| AMD RX 7900 XTX 24GB | 93.5 |
| RTX 3090 24GB | 86.6 |
| RTX 3090 Ti 24GB | 85.3 |
| AMD RX 6900 XT 16GB | 82.4 |
| RTX 4090 24GB | 81.8 |
| RTX 3080 Ti 12GB | 79.7 |
| RTX 4080 16GB | 77.9 |
| RTX 2080 Ti 11GB | 75.6 |
| RTX 4070 Ti 12GB | 71.7 |
The data allows us to draw a few conclusions and make some interesting observations.
Firstly, AMD’s RX 7900 XTX 24GB tops the list with a score of 93.5, outperforming all other GPUs in this comparison. This suggests that AMD has made significant strides in GPU technology, and the RX 7900 XTX has managed to claim the performance crown in this particular evaluation.
NVIDIA’s RTX 3090 24GB and RTX 3090 Ti 24GB follow closely with scores of 86.6 and 85.3 respectively. These are still impressive scores, indicating that NVIDIA’s GPUs offer strong performance as well. However, it is interesting to note that the RTX 3090 Ti, a newer and presumably more advanced model than the RTX 3090, has a slightly lower score than its predecessor. This may be due to factors such as silicon binning, manufacturing variations, or drivers.
The mid-range GPUs in the data are represented by the AMD RX 6900 XT 16GB, RTX 4090 24GB, RTX 3080 Ti 12GB, and RTX 4080 16GB. These GPUs score between 77.9 and 82.4, offering solid performance for their respective price brackets.
Extended Overall Score
| GPU | Score |
| RTX 4090 24GB | 2960 |
| RTX 4080 16GB | 2768 |
| AMD RX 7900 XTX 24GB | 2745 |
| RTX 3090 Ti 24GB | 2728 |
| RTX 3090 24GB | 2718 |
| RTX 3080 Ti 12GB | 2583 |
| RTX 4070 Ti 12GB | 2483 |
| RTX 3080 10GB | 2448 |
In this comparison, NVIDIA’s RTX 4090 24GB leads the pack with a score of 2960, showcasing its strong performance and positioning it as the top choice for high-end gaming and demanding tasks. The RTX 4080 16GB follows closely with a score of 2768, highlighting its impressive capabilities and making it a solid contender in the high-performance GPU segment.
Interestingly, the AMD RX 7900 XTX 24GB, which topped the list in the previous data set, ranks third in this evaluation with a score of 2745. While it still offers robust performance, it falls slightly behind the top NVIDIA GPUs in this Extended Overall Score comparison.
In the lower spectrum range, the RTX 4070 Ti 12GB and RTX 3080 10GB score 2483 and 2448, respectively. Although these GPUs may not offer top-tier performance, they still provide a satisfactory experience for less demanding tasks or gaming in lower settings, making them a reasonable choice for budget-conscious consumers.
GPU Effects Score
| GPU | Score |
| RTX 4090 24GB | 258 |
| RTX 3090 Ti 24GB | 191 |
| RTX 4080 16GB | 189 |
| AMD RX 7900 XTX 24GB | 182 |
| RTX 3090 24 GB | 181 |
| RTX 3080 Ti 24GB | 169 |
| RTX 3080 10 GB | 152 |
| RTX 4070 Ti 12GB | 145 |
| AMD RX 6900 XT 16GB | 145 |
NVIDIA’s RTX 4090 24GB once again takes the lead with a dominant score of 258, showcasing its exceptional capability in handling advanced graphical effects. As a result, this GPU would be an ideal choice for video editing professionals who demand top-tier performance for handling complex projects.
The RTX 3090 Ti 24GB and RTX 4080 16GB follow with scores of 191 and 189, respectively. Again, these GPUs demonstrate strong performance in rendering graphical effects, making them solid options for users who seek high-quality visuals in their gaming or professional work.
The AMD RX 7900 XTX 24GB scores 182, which places it slightly below the top NVIDIA GPUs in terms of GPU effects capabilities. Despite this, it still offers robust performance and would be a viable choice for users prioritizing a blend of performance and value.
Click here to check the full analysis.