Puget Systems print logo
https://www.pugetsystems.com
Read this article at https://www.pugetsystems.com/guides/2033
Article Thumbnail

Agisoft Metashape 1.6.5 - GeForce RTX 30 Series vs Radeon RX 6000 Series

Written on January 6, 2021 by William George
Share:

TL;DR: RTX 30 Series vs RX 6000 Series Performance in Metashape

Any modern video card will speed up processing in this application to some degree, but NVIDIA's GeForce RTX 30 Series cards are clearly faster than AMD's Radeon RTX 6000 Series in Metashape 1.6.5. The bigger the project is, the more this difference will be noticeable. And importantly, dual GeForce cards can offer even better performance than a lone card for less money.

Introduction

Both NVIDIA and AMD launched new lines of video cards in late 2020, and the various models from both companies have trickled out over time since the original announcements. We have finally gotten samples of the most recently released models, the GeForce RTX 3060 Ti and Radeon RX 6900 XT, so this is a great time for a round-up of modern GPU performance in Agisoft Metashape.

Agisoft Metashape Photogrammetry Performance with GeForce RTX 30 Series vs AMD Radeon RX 6000 Series

If you want to see the full specifications for these new video cards, we recommend checking out the AMD and NVIDIA product pages. But at a glance, here are what we consider to be the most important specs:

VRAM Cores Boost Clock Power MSRP
GeForce RTX 3060 Ti 8GB 4864 1.67 GHz 200W $399
GeForce RTX 3070 8GB 5888 1.70 GHz 220W $499
Radeon RX 6800 16GB 3840 2.10 GHz 250W $579
Radeon RX 6800 XT 16GB 4608 2.25 GHz 300W $649
GeForce RTX 3080 10GB 8704 1.71 GHz 320W $699
Radeon RX 6900 XT 16GB 5120 2.25 GHz 300W $999
GeForce RTX 3090 24GB 10496 1.73 GHz 350W $1499

While specs rarely line up with real-world performance, it is nice to see AMD including 16GB of VRAM on their new Radeon 6000 Series cards. However, based on our past testing, this really shouldn't affect Metashape performance in any meaningful way.

Metashape Workstations

Puget Systems offers a range of powerful and reliable systems that are tailor-made for your unique workflow.

Configure a System!

Labs Consultation Service

Our Labs team is available to provide in-depth hardware recommendations based on your workflow.

Find Out More!

Test Setup

Here are the specifications of the system we used for our Metashape testing, with components selected to minimize other bottlenecks:

* Please note that we ran AMD's Ryzen 9 5950X processor at stock settings for this testing, but we have found that even better performance is available from this CPU (and other high core count models) by turning off SMT.

To test each video card, we used our in-house Metashape benchmark - both the standard and extended versions. Each was run twice per GPU, and the best results were included in the charts below. Here is some basic info about the image sets in these projects:

  • Rock Model - 36 photos at 20 megapixels each
  • School Map - 51 photos at 18 megapixels each
  • School Model - 439 photos at 18 megapixels each
  • Park Map - 792 photos at 18 megapixels each

We make these benchmarks publicly available under the Creative Commons BY-ND license, so if you use Metashape Professional (or get the 30-day trial) you can download them from our website and compare your system's performance with the results shown here.

Benchmark Results

Here are charts showing the overall per-project performance of the AMD Radeon RX 6000 cards (in blue) compared to NVIDIA's competing GeForce RTX 30 Series (green) and no active GPU (grey):

Performance Analysis

The overall spread in our smaller Metashape image sets isn't huge, with no more than a few percent variance there between all seven cards we tested, but if you are working with larger projects (hundreds of images or more) then the differences between GPUs become very pronounced. And the delta isn't so much between models within the same product family, though that is measurable, but rather between AMD's Radeon RX 6000 line and NVIDIA's GeForce RTX 30 Series. In both of our larger image sets, the AMD cards were 50-70% slower than their NVIDIA counterparts - which is a massive drop in performance, unfortunately.

But where is such a big difference coming from? To answer that, we have to look at the how these cards fare in each processing step:

Metashape 1.6.5 GeForce RTX 30 Series vs Radeon RX 6000 Series Performance Table

Table showing processing time per step (lower is better)

Pouring over the table above provides a few interesting insights:

  1. The AMD Radeon cards are as fast or faster than NVIDIA's GeForce when it comes to Align Photos, Build Dense Cloud, Build Mesh, and Build Texture steps. They are just a hair behind in Build Depth Maps, but not enough to be a big concern.
  2. In Map projects, both the Build DEM and Build Orthomosaic steps appear to be CPU only (since they perform the same across all the video cards as well when no GPU is selected in Preferences). Build Dense Cloud may fit this category as well, but it does seem to vary a little between the AMD and NVIDIA cards.
  3. Decimate Mesh is, oddly, slower with AMD cards than with no GPU selected at all (and faster on NVIDIA cards). It is an extremely short part of the overall photogrammetry process, however, so it isn't worth worrying about.
  4. The big performance issue with AMD cards, then, comes from a single step: Build Depth Maps. Having one of these Radeon cards does reduce processing time by 5-10% in this step on our large image sets, compared to no GPU acceleration, but having a NVIDIA GeForce card instead drops that time by ~80%! Since this is also one of the longest steps in Metashape's workflow, the massive lead NVIDIA cards have here carries them to a huge victory in overall processing time as well.

Here is a chart showing processing time for that Build Depth Maps step on the Park Map project, to help visualize this difference:

GeForce RTX 30 Series vs Radeon RX 6000 Series Processing Times for Build Depth Maps Step in Metashape 1.6.5

Is NVIDIA's GeForce RTX 30 Series Good for Agisoft Metashape?

Yes - all of the GeForce RTX 30 Series cards do quite well in Metashape, and there isn't a huge performance spread so even the more afforable RTX 3070 and 3060 Ti are solid choices for this application. Moreover, we have found that pairing up two of these cards increases processing speed even further, by 5-10%, such that dual RTX 3070 cards will outperform a single RTX 3090 for less money!

Is AMD's Radeon RX 6000 Series Good for Agisoft Metashape?

Unfortunately, no - at least as of version 1.6.5, these new AMD Radeon cards are not able to keep up with NVIDIA's competing GeForce RTX 30 Series video cards in Metashape. Better optimization of Metashape for AMD cards, especially in the Build Depth Maps step, might be able to improve their performance in the future - but for now we would recommend sticking with NVIDIA for this application.

As always, please keep in mind that these results are strictly for photogrammetry in Agisoft Metashape. If you have performance concerns for other software in your workflow, we highly recommend checking out our Hardware Articles (you can filter by "Video Card") for the latest information on how a wide range of programs perform with various GPUs, CPUs, and other hardware.

Photogrammetry Workstations

Puget Systems offers a range of poweful and reliable systems that are tailor-made for your unique workflow.

Configure a System!

Labs Consultation Service

Our Labs team is available to provide in-depth hardware recommendations based on your workflow.

Find Out More!
Tags: Hardware Acceleration, NVIDIA, RTX 3080, GPU Acceleration, GPU, Video Card, Metashape, Agisoft, photogrammetry, RTX 3070, AMD, Radeon, Radeon RX 6800, Radeon RX 6800 XT, GeForce, Radeon RX 6900 XT, RTX 3060 Ti, RTX 3090
JoshWAL

What's the point of constantly posting the same useless & buggy results ?!
The depth maps are clearly not being properly generated by the RX6XXX GPUs for the School & Park Map projects. It's either an AMD driver issue (only the Radeon Pro 20.Q4 drivers have a non broken OpenCL support right now), Metashape 1.6.5 or your Benchmark.
Your very own RX580/ Vega 64 / RadeonVII benchmarks proves that. The Vega 64 & RadeonVII are literally x3.6 faster :
https://www.pugetsystems.co...
https://uploads.disquscdn.c...

Posted on 2021-01-08 02:38:43

We had results for the newer cards (3060 Ti and 6900 XT) - why wouldn't we post them?

Those older AMD cards you mentioned used a different architecture, so I wouldn't use that older article to analyze these newer GPUs.

However, I do have some good news! I started testing Metashape 1.7 (which came out on New Years Day, after we had finished most of our testing for the article above) and it looks like AMD cards are performing a lot better! I hope to have a post about that next week, but need to test a few more cards still.

Posted on 2021-01-08 03:00:02
JoshWAL

"We had results for the newer cards (3060 Ti and 6900 XT) - why wouldn't we post them?"

Because it is obviously clear that those results are wrong because of either a driver/soft/project issue and you simply post them without any warning or even a link back to your own previous benchmark which shows 3 years old GPUs being x3.6 faster?
(Even if those GPUs where GCN Uarch the compute units are nearly identical on RDNA1/2 and the are not going to magical be x3.6+ slower come on

Posted on 2021-01-08 03:23:26

Over the years, some of the parameters available in Metashape's Python scripting have changed - so I wouldn't directly compare results using different versions of our benchmark like that. But as I said in the conclusion to this article: "Better optimization of Metashape for AMD cards, especially in the Build Depth Maps step, might be able to improve their performance in the future" - and lo and behold, it has happened! Check out the new article, looking at Metashape 1.7.0 :)

https://www.pugetsystems.co...

Posted on 2021-01-09 00:33:19
Michael Freeman

What can we expect with the 3080Ti? Will it be the same as the 3090?

Posted on 2021-09-20 17:26:48

I don't have data on the 3080 Ti in Metashape specifically, but it generally falls between the 3080 and 3090 in terms of performance. Those cards are so close in Metashape already that I think it should be a solid option - but maybe not worth the added price over the 3080 for a lot of users.

Posted on 2021-09-20 17:28:37