Table of Contents
After Effects has had a bit of a rocky relationship with video cards ever since GPU acceleration was added back in 2015. At that time, After Effects dropped support for the "render multiple frames simultaneously" (which allowed for efficient use of high core count CPUs) in part to make way for GPU acceleration. While obviously frustrating for users who had invested thousands of dollars in dual Xeon workstations, for most users this resulted in great performance improvements at little to no cost. But even with the improvements made to AE's GPU acceleration over the years, there has been little reason to use more than a mid-range video card.
However, NVIDIA's new RTX series cards are here and they bring to the table two new features that may finally give you a reason to invest in a high-end GPU for After Effects: Tensor cores and RT cores.
What are Tensor Cores?
While already available on the more expensive Titan V GPU, the RTX line introduces tensor cores at a more reasonable price point. These tensor cores operate alongside the normal CUDA cores that traditionally do the heavy lifting, but are designed specifically for machine learning inference (running already created and trained machine learning models). While not utilized in Adobe applications today, these cores may be used in the future – especially as Adobe continues to develop their "Adobe Sensei" AI and machine learning technology.
What are RT Cores?
RT cores are brand new in this generation of graphics cards, and are specialized for a single type of operation: ray tracing. It is possible that Adobe may utilize these cores for ray tracing in After Effects – or Maxon may use it in Cinema 4D – but if or when they will take advantage of these RT cores is currently unknown.
What makes these new RTX cards hard to review and test is the fact that After Effects currently does not use either of these new types of cores. We can (and will) look at straight performance gains with the current version of After Effects, but really what you are paying for is technology that might give you significant performance gains in the future.
If you would like to skip over our test setup and benchmark result/analysis sections, feel free to jump right to the Conclusion section.
Test Hardware & Methodology
Listed below is the test platform we will be using in our testing:
|Gigabyte Z370 Aorus Gaming 5
|Intel Core i7 8700K 6 Core
3.7GHz (4.7GHz Turbo)
|4x DDR4-2666 16GB (64GB total)
|Samsung 960 Pro 1TB M.2 PCI-E x4 NVMe SSD
|Windows 10 Pro 64-bit
|After Effects CC 2018 (ver. 15.1.2)
To see how the new RTX cards perform in After Effects, we tested it against a selection of cards from NVIDIA as well as AMD's Vega 64 GPU.
In order to accurately benchmark the different systems, we used a range of After Effects projects that are mostly publicly available for download. The projects we used (along with their source) are:
|0-40 (40 total frames)
by Video Copilot
|212-238 (26 total frames)
0-97 (97 total frames)
by Video Copilot
|0-78 (78 total frames)
|32610-32710 (100 total frames)
4K H.264 video with:
|0-169 (169 total frames)
"Video Wall" and "*.mov" layers
RAM Preview – Raw Benchmark Data
RAM Preview – Benchmark Analysis
In After Effects, there is a big difference between standard projects and those that utilize the Cinema 4D CPU renderer. This mostly impacts CPU performance, but since there is the chance that it will impact our GPU performance results as well, we decided to separate out our testing results between "standard" projects and those utilizing the C4D CPU renderer.
Before we get into the results themselves, we want to explain the scoring system used in this test. In essence, a score of "20" would mean that on average that processor was able to play our projects at 20% of the project's defined FPS. A perfect score would be "100" which would mean that the system was able to play it back in real time, although with the difficult projects we use this should never actually occur.
As we expected, since the C4D rendering engine relies almost exclusively on the CPU there was almost no difference in performance between each GPU for those projects. For the standard projects, however, there is a small benefit to using a more powerful GPU, but it isn't very much. Here, we saw a small gain in performance up to about the GTX 1080 Ti – after which the results were pretty much within the margin of error for this test. This means that we only saw a whopping 5% performance gain with the new RTX cards over the much more modest GTX 1060 GPU.
Final Render – Raw Benchmark Data
Final Render – Benchmark Analysis
Once again, since the results may vary based on whether or not you utilize the Cinema 4D CPU renderer, we have separated our testing results between "standard" projects and those utilizing the C4D renderer:
For the final render (or export), we saw a bit larger of a performance gap between each of the cards. The Vega 64 GPU also surprised us a bit here, performing right in line with the top-end NVIDIA cards. Once again, however, we seem to top out at about the GTX 1080 Ti and above since the GTX 1080 Ti, Titan Xp, Titan V, and the new RTX cards all performed about 10% faster than the GTX 1060.
Are the RTX video cards good for After Effects?
GPU acceleration is relatively new to After Effects, and it still shows in how well (or not well) it is able to utilize high-end GPUs. While the RTX 2080 and 2080 Ti are certainly not bad for AE, they probably aren't worth purchasing for their raw performance alone.
The score in the chart above is a weighted average of our testing based on what our customers tend to be the most concerned about. RAM Preview of standard projects makes up 40% of the overall score while the Ram Preview (C4D Renderer), Final Render (Standard), and Final Render (C4D Renderer) tests each contribute 20% towards the score.
In terms of pure performance, the new RTX 2080 and 2080 Ti perform on par with the GTX 1080 Ti and slightly under the Titan Xp and Titan V. However, the question is not really whether they are good today, but whether the RTX GPUs will be good for After Effects in the future. The new RT and tensor cores could potentially have significant performance advantages for some tasks, but it all depends on when or if the After Effects developers are able to make effectively use of these new features. Given how long it has taken for GPU acceleration to make it into After Effects in the first place, our guess is that it will be quite a while before we see significant performance gains from the RT and tensor cores, but there is really no way to know for sure.
Overall, while not really necessary for After Effects alone, if you already need a new high-end GPU for other applications (Premiere Pro, GPU-based rendering, etc.) we would recommend using one of these RTX cards if possible. The RTX 2080 does have less VRAM than the comparably priced GTX 1080 Ti (8GB vs 11GB), but the potential these cards offer for the future is likely worth investing in.
If you are interested in how the RTX cards perform in other applications, be sure to check out our recent Video Card articles as we have (or are working on) a number of other articles for the RTX 2080 and RTX 2080 Ti.
Puget Systems offers a range of powerful and reliable systems that are tailor-made for your unique workflow.