Puget Systems print logo

https://www.pugetsystems.com

Read this article at https://www.pugetsystems.com/guides/1310
Article Thumbnail

DaVinci Resolve 15 CPU Roundup: Intel vs AMD vs Mac

Written on December 10, 2018 by Matt Bach
Share:

Introduction

Unlike most other applications used in the video post-production process, DaVinci Resolve relies much more on the power of your video card(s) than it does on your processor. However, you do need enough CPU power to keep up with your GPU(s) and there are still some aspects of Resolve (like the Fusion tab) where the CPU is more important.

In this article, we will be looking at a wide range of processors from Intel and AMD including the Intel 9th Gen, Intel X-series, AMD Ryzen 2nd Gen, and AMD Threadripper 2nd Gen CPU lines. In addition, we will be comparing them to a current Mac Pro 12 Core and iMac Pro 14 Core for those that are curious about how much faster a PC workstation can be compared to a Mac.

One thing to note is that we will not be including results for any previous-gen CPUs in this article. At first, we were going to include them but the charts and tables soon got out of hand. Instead, if you want to know how these CPUs compare to previous generations, we recommend checking out the following articles:

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 Setup & Methodology

Listed below are the systems we will be using in our testing:

Shared PC Hardware/Software
Video Card: 1-2x NVIDIA GeForce RTX 2080 Ti 11GB
Hard Drive: Samsung 960 Pro 1TB M.2 PCI-E x4 NVMe SSD
OS: Windows 10 Pro 64-bit (version 1803)
Mac-based PC Test Hardware  
System: Apple Mac Pro (12 Core) Apple iMac Pro (14 Core)
CPU: 12-core
2.7GHz 30MB of L3 cache
14-core Intel Xeon W
2.5GHz Turbo Boost up to 4.3GHz
RAM: 64GB 1866MHz DDR3 ECC 64GB 2666MHz DDR4 ECC
Video Card: Dual AMD FirePro D700, 6GB of GDDR5 VRAM Radeon Pro Vega 64, 16GB of HBM2 memory
Hard Drive: 1TB PCIe-based SSD 1TB SSD
OS: MacOS Mojave (10.14.1)

Our testing was done on DaVinci Resolve Studio 15.1.2.8 and primarily revolves around the Color tab and focuses on the minimum FPS you would see with various media and levels of grading. The lowest level of grading we test is simply a basic correction using the color wheels plus 4 Power Window nodes with motion tracking. The next level up is the same adjustments but with the addition of 3 OpenFX nodes: Lens Flare, Tilt-Shift Blur, and Sharpen. The final level has all of the previous nodes plus one TNR node.

We kept our project timelines at Ultra HD (3840x2160) across all the tests, but changed the playback framerate to match the FPS of the media. For all the difficult RAW footage we tested (CinemaDNG & RED), we not only tested with the RAW decode quality set to "Full Res" but we also tested at "Half Res" ("Half Res Good" for the RED footage). Full resolution decoding should show the largest performance delta between the different cards, but we also want to see what kind of FPS increase you might see by running at a lower decode resolution.

Codec Resolution FPS Bitrate Clip Name Source
H.264 3840x2160 29.97 FPS 80 Mbps Transcoded from RED 4K clip
H.264 LongGOP 3840x2160 29.97 FPS 150 Mbps Provided by Neil Purcell - www.neilpurcell.com
DNxHR HQ 8-bit 3840x2160 29.97 FPS 870 Mbps Transcoded from RED 4K clip
ProRes 422 HQ 3840x2160 29.97 FPS 900 Mbps Transcoded from RED 4K clip
ProRes 4444 3840x2160 29.97 FPS 1,200 Mbps Transcoded from RED 4K clip
XAVC S 3840x2160 29.97 FPS 90 Mbps Provided by Samuel Neff - www.neffvisuals.com
XAVC Long GOP 3840x2160 29.97 FPS 190 Mbps Transcoded from RED 4K clip
Blackmagic RAW 4608x1920 24 FPS 210 Mbps A001_08122231_C008 Blackmagic RAW
RED (7:1) 4096x2304 29.97 FPS 300 Mbps A004_C186_011278_001 RED Sample R3D Files
CinemaDNG 4608x2592 24 FPS 1,900 Mbps Interior Office Blackmagic Design
[Direct Download]
RED (7:1) 6144x3077 23.976 FPS 840 Mbps S005_L001_0220LI_001 RED Sample R3D Files
RED (9:1) 8192x4320 25 FPS 1,000 Mbps B001_C096_0902AP_001 RED Sample R3D Files

With the addition of the "Fusion" tab in Resolve, we are also going to be including some basic tests for that tab as well. At the moment these are relatively easy projects that specifically test things like particles with a turbulence node, planar tracking, compositing, and 3D text with a heavy gaussian blur node. These projects are based on the following tutorials:

If you have suggestions on what we should test in the future, please let us know in the comments section. Especially if you are able to send us a sample project to use, we really want to hear from you!

Benchmark Results

While our benchmark presents various scores based on the performance of each type of task, we also wanted to provide the individual results in case there is a specific task someone may be interested in. Feel free to skip to the next section for our analysis of these results.

Color Tab FPS - Benchmark Analysis

To analyze our benchmark results, we are going to break it down based on whether we used one or two RTX 2080 Ti GPUs and the three different levels of color grading we tested. The easiest level of grading - a basic grade with 4 power windows - is not too difficult and every CPU we tested should be able to give full playback FPS with everything but RED 6K/8K media. However, each level up should show more and more of a difference between the different processors.

The "Score" shown in the charts is a representation of the average performance we saw with each GPU for that test. In essence, a score of "80" means that on average, the system was able to play our project at 80% of the tested media's FPS. A perfect score would be "100" which would mean that the system gave full FPS even with the most difficult codecs and grades.

DaVinci Resolve Studio 15 Single GPU Benchmark CPU Roundup - Intel 9th Gen, Intel X-series, AMD Threadripper 2nd Gen, AMD Ryzen 2nd Gen, Apple Mac Pro, Apple iMac Pro

Starting with the Color Grading results with a single RTX 2080 Ti, you can really clearly see that while you need to have enough CPU power to keep up with your GPU(s), there isn't much benefit to having more CPU power than you need. In fact, all the AMD Threadripper CPUs and all the Intel X-series CPU (with the exception of the i7 9800X) pretty much performed the same. So, if you wanted to get the full performance out of your RTX 2080 Ti, as long as you simply use an Intel X-series i9 or AMD Threadripper CPU, you should be just fine.

If those CPUs are outside your budget, the Intel 9th Gen CPUs also did pretty well, although at this level the AMD Ryzen 2700X is likely a better choice since it is typically priced closer to the i5 9600K while being ~30% faster for color grading in Resolve.

On the Mac side, the iMac Pro and Mac Pro don't use particularly powerful GPUs which means that they are not ideal for high-end color grading work in DaVinci Resolve. Even the iMac Pro with the Vega 64 GPU was only able to beat the Core i7 9600K, and even then only by ~8%. Keep in mind that this isn't a completely fair comparison in terms of CPU power alone since we would need to be using an AMD Radeon Vega 64 on the PC in order to get a direct comparison.

DaVinci Resolve Studio 15 Dual GPU Benchmark CPU Roundup - Intel 9th Gen, Intel X-series, AMD Threadripper 2nd Gen, AMD Ryzen 2nd Gen, Apple Mac Pro, Apple iMac Pro

With dual RTX 2080 Ti, we see a bit more of a benefit to using a more powerful CPU. With this setup, we would recommend at least the Intel X-series i9 9920X 12 Core or the AMD Threadripper 2950X 16 Core in order to get full performance. You can go even higher-end if you want to of course, but there is almost no performance reason to do so unless you are planning on adding a third or fourth high-end GPU to the system.

Fusion Tab FPS - Benchmark Analysis

DaVinci Resolve Studio 15 Fusion Benchmark CPU Roundup - Intel 9th Gen, Intel X-series, AMD Threadripper 2nd Gen, AMD Ryzen 2nd Gen, Apple Mac Pro, Apple iMac Pro

Moving on to the results for Fusion, the first thing we want to point out is that having multiple GPUs doesn't appear to greatly affect performance. To be fair, we are not using footage inside any of our projects that is particularly difficult to process, but given the FPS we saw in each project we doubt that that having multiple GPUs would significantly improve performance even if you are using 8K RED media. Because of this, we are going to focus on the results we saw with a single RTX 2080 Ti video card.

Fusion is very different than the rest of Resolve since it is more about single-threaded CPU performance than anything else. In fact, these results are similar to what we saw for After Effects. In short, the Intel 9th Gen CPUs are king, easily beating the Intel X-series and AMD Threadripper CPUs. The one bit of a surprise to us was that the 14 core iMac Pro was able to keep up very well with the Intel 9th Gen CPUs in the Fusion Tab. It wasn't quite as fast as the i9 9900K or i7 9700K, but it certainly did better than the Intel X-series or AMD Threadripper CPUs.

Something interesting to point out is the low performance we saw with the AMD Threadripper "WX" models. Those models are essentially two CPUs crammed onto a single die which results in additional overhead that can cause performance loss in applications that cannot take advantage of higher core counts. Even with the new "Dynamic Local Mode" feature enabled (which is supposed to assist in this sort of workload) the WX models clearly have some performance issues.

Intel vs AMD vs Mac for DaVinci Resolve 15

In DaVinci Resolve, performance is determined more by the power of your GPU(s) than your CPU, but it is still important to have enough CPU power to keep up. In most cases, we would recommend using an Intel X-series i9 CPU or an AMD Threadripper CPU in order to ensure that you are getting the best performance for your dollar. For a mid-range CPU (roughly $500-$1200) Threadripper should give you slightly better performance for color grading work, although at higher price points the Intel X-series take the lead.

Where it really gets complicated is if you use (or plan to use) the relatively new Fusion tab. Fusion is much more about single threaded CPU performance, so CPUs like the Intel 9th Gen i9 9900K are king. The Intel X-series still do pretty well in Fusion - which makes them a great balance between the Color and Fusion Tabs - although AMD Threadripper CPUs tend to underperform slightly since their per-core performance is currently lower than Intel.

Our testing included the Mac Pro and iMac Pro, but to be honest this isn't a very fair CPU comparison. Since the Mac systems use AMD graphics with relatively mid-range GPU performance, our testing doesn't really show what the Mac systems can do on a pure CPU level. We could have used a Radeon Vega 64 in all our testing, but for the price of a 14-core iMac Pro you can easily get a PC with dual RTX 2080 Ti GPUs which should be about 40% faster in DaVinci Resolve than the iMac Pro. So, while this isn't fair to the CPU alone, it still is very informative on how a Mac system could compare to a PC in DaVinci Resolve.

If you are curious how the latest Intel and AMD processors perform in other applications, be sure to check out our recent Processor articles as we have a number of articles looking at CPU performance in Photoshop, Lightroom, After Effects, Premiere Pro, and many other software packages.

Tags: DaVinci Resolve, Intel 9th Gen, Intel X-series, AMD Threadripper 2nd Gen, AMD Ryzen 2nd Gen, Apple Mac Pro, Apple iMac Pro, 9900K, 9700K, 9600K, 9980XE, 9960X, 9940X, 9920X, 9900X, 9980X, 9800X, 2990WX, 2970WX, 2950X, 2920X, 2700X
Cameron Porter

Fascinating read. Thanks for putting this together. Why do applications like Fusion and After Effects rely so heavily on single-core performance? Is that simply the nature of precise VFX rendering or is that something that could be addressed by developers in the future?

Posted on 2018-12-11 16:45:10

From what I understand, it is really just the nature of the kind of calculations going on in the background. If the program has to do 20 calculations, but each relies on the results from the previous calculation, you can't really do that in parallel. Not everything in Ae/Fusion is single threaded of course - color correction is a prime example of something you can do that should be highly parallel - but many of those tasks are actually moved onto the GPU since video cards are way better at processing highly parallel tasks than CPUs are. So what you end up with is all the parallel tasks being done on the GPU which leaves just the single/lightly threaded tasks left on the CPU.

Or it could just be that the developers haven't done a good job making their code run in parallel when it could be. I honestly don't think that is the case for Ae and Fusion (although I'm sure there are plenty of improvements that could be made), but that is a pretty common issue in software with smaller or less experienced dev teams.

Posted on 2018-12-11 17:24:27
Turing

https://forum.blackmagicdes...

"We are pleased to announce the release of DaVinci Resolve 15.2.2."

Posted on 2018-12-13 10:16:29
old germanian komputar

Use Thread-ripper 12 core 2gen and i9 10 core ??gen? ( hackintosh). The marginal diff is that both sys have vega 56 ( and lower heat/consumption) no Nvidia ( c4d is amd-fan)( texture-progs are open dogs/ z-brush likes your gpu...and works proper in both) and are in this case a sweet spot for my moneybags vs render-power. Both run with 64 gbt ddr-4 and both run on normal ssd bx-500 ( 4tbt).. m2-960 ( 1tbt) for the os and progs. osx redhat w7 unix-normallo asshole-w-10
( gpu-rendering in redshift is now a bit out= since majestic nvida will bring gtxrtgtt 3000-series GPU, under 499,- / 202?)-

Davinic is the lowest consumption Factory, if we go in 10 bit 4k normal video-signal. What goes fine in midprice film/ vfx [ bit upper sharknado ] :)
Adobeeeeee.... a big party of nooge and dope and it is a hunger game, and like AUTODESK the MONTHLY-WHORE_THAT_DOES_NOT_SUCK_DEEP!!
a shithose..and now a stabel crak KRAGG!
Final Cut.. small show
Natron 39% of big british mama Nuke, if you paint in mari (aaaa..) and if Osx-last december = hakintosh is fine i5 i makk pro system shut down..

it is a funny game, but the imac will ne last one. after 22 years, Apple lost me. no power expensive and old.. it is like staying in 2009.
no :)

Posted on 2018-12-17 15:07:52
Charles Unice

I’m not sure I understand your graph. In the explanation you say a score of 100 means it plays 100% of the media. Then your graph gives score of 944 for dual GPU th 1950x does that mean it plays 944% of your media? That doesn’t make sense to me. Sorry I’m Seriously just trying to understand your data! You know I love you guys.

Posted on 2019-01-06 21:32:35

That is for the individual test charts. For the overall, it is a combination of each color grading test so we increase the threshold by 10x in order to differentiate it from the individual sections. Pretty much the same thing, just a score of 944 is 94.4% of full FPS on average across all the tests.

Posted on 2019-01-07 17:50:44