Scaling Motion Design in Cinema 4D and Redshift: A Multi-GPU Workflow Guide

Motion design jobs combine everything that makes rendering slow: thousands of MoGraph clones, glass and volumetric lighting, simulations, several aspect ratios and tight revision cycles. Moving such a project to a multi-GPU machine helps only if the scene is set up to use it.

This guide covers how Redshift uses several GPUs, how to keep MoGraph scenes light with instancing and proxies, how to make simulations and output paths safe for any machine, and how to split a delivery across servers. It follows Maxon’s Cinema 4D and Redshift documentation.

In Bucket mode, which Maxon recommends for final renders, Redshift splits each frame into buckets. Maxon’s documentation explains that a bucket is only ever rendered by a single GPU, so the number of buckets rendered at the same time is limited by the number of GPUs Redshift is set to use. More GPUs therefore render more buckets in parallel on the same frame.

Two things limit how far that goes. First, every GPU holds its own copy of the scene, because memory is not added up across cards. Second, Cinema 4D has to prepare the scene, evaluating cloners, effectors, deformers and simulations, before any bucket starts, and that work does not get faster with more GPUs. In iRender’s own tests, a 4x RTX 5090 server renders Redshift about 3.7 times as fast as one card, and an 8x RTX 5090 server up to 7.2 times; heavy scene preparation lowers those numbers, which is why the next sections matter.

Maxon’s Redshift documentation says Redshift automatically converts Cinema 4D instances, such as Instance objects and Cloners, into equivalent Redshift instances where supported; objects Redshift cannot instance, for example lights and volumes, are translated as copies. How much work that saves depends on the Cloner’s Instance Mode:

For a full walkthrough, including per-clone colours with Multi-Instance, see our guide to MoGraph clones in Redshift.

Trees, crowds, vehicles and set dressing that repeat across shots are good candidates for Redshift Proxies (.rs files). Cinema 4D keeps only a lightweight placeholder, and Redshift loads the contents when it needs them. Export with File > Export > Redshift Proxy, place them with a Redshift Proxy object, and clone the proxy object with a Cloner in Render Instance or Multi-Instance mode. Keep the .rs files and their textures inside the project folder so another machine finds them. Our Redshift Proxy guide covers export options and texture paths.

Dynamics, cloth, particles and MoGraph effects that depend on previous frames should be cached before rendering, with the caches stored inside the project. An uncached simulation is recalculated on the render machine, which adds time to every frame, and when frames are split between machines each one starts part-way through the simulation, so the result can differ from what you approved.

Motion design jobs often ship several versions of the same shot, for example 16:9 and 9:16. Cinema 4D Takes can hold the camera, resolution and render settings for each version. Use tokens in the output path, such as $prj, $take, $camera and $frame, so every take writes to its own file names and nothing is overwritten. Before rendering, check which takes are marked for rendering.

Within one frame, more GPUs render more buckets at once. Across a sequence, the simplest way to finish sooner is to give each server its own frame range: three 4-GPU servers each rendering a third of the frames finish in about a third of the time for about the same total cost. This works best when simulations are cached and every server reads the same project folder.

Both farm models render the same frames from the same project. The difference for motion design is how revision rounds work: change a clone setup or a light, look at the result, change it again.

A simple rule: if the shot is locked and you only need frames, a SaaS farm is a good fit. If revisions are still coming, or you need to test instancing, proxies and caches on the render hardware, use an IaaS server.

1. Save Project with Assets, with caches and proxies inside the project folder, and upload it to iRender Drive before you start the server; uploading is free and does not use rental time.
2. Start a 4x or 8x RTX 5090 server. Cinema 4D and Redshift are installed. Billing is per minute.
3. Render the heaviest frame of each take in Bucket mode and note the scene preparation time and the render time.
4. Make revisions on the server and render the changed frames straight away.
5. Render the deliverables. If the deadline is tight, start more servers and give each one a frame range or a take; the total cost stays about the same.

• Use the same Cinema 4D release and Redshift build.
• Set Cloners to Render Instance or Multi-Instance where you can, and use Redshift Proxies for heavy repeated assets.
• Cache simulations and keep caches, proxies and textures in the project folder.
• Check takes, cameras and output-path tokens for every deliverable.
• Render the heaviest frame of each take in Bucket mode before the full range.

1. How does Redshift use multiple GPUs on one frame?

In Bucket mode Redshift splits the frame into buckets, and Maxon’s documentation says each bucket is only ever rendered by a single GPU. More GPUs therefore render more buckets at the same time, while each GPU keeps its own copy of the scene in its own memory.

2. Which Cloner Instance Mode should I use for Redshift?

Use Multi-Instance for very large clone counts and Render Instance for most other cloner setups, or when a feature does not work in Multi-Instance. Plain Instance creates a full copy of every clone and only suits small counts. Redshift converts supported Cinema 4D instances into its own instances.

3. When should I use Redshift Proxies in motion design?

Use them for heavy assets that repeat, such as trees, crowds, vehicles or set dressing. Cinema 4D keeps only a placeholder and Redshift loads the contents when it needs them, which keeps the scene light. Keep the .rs files and their textures in the project folder.

4. Why does my simulation look different when frames are split across machines?

If the simulation is not cached, each machine recalculates it, and a machine that starts part-way through the sequence may not reach the same state. Cache dynamics, cloth, particles and MoGraph effects before rendering and keep the caches in the project folder.

5. How much faster is Redshift on 4 or 8 GPUs?

In iRender’s own tests a 4x RTX 5090 server renders Redshift about 3.7 times as fast as one card, and an 8x RTX 5090 server up to 7.2 times. Heavy scene preparation in Cinema 4D lowers the gain, so measure your heaviest frame.

6. Can I make revisions on the machine that renders the motion design job?

Yes, on an IaaS render farm. On iRender you rent the whole server with Cinema 4D and Redshift installed, so you can change the scene, render the changed frames straight away and then render the deliverables on the same GPUs.

iRender is an IaaS GPU render farm: you rent a bare-metal server with 1, 2, 4 or 8 NVIDIA RTX 4090 (24 GB) or RTX 5090 (32 GB) GPUs, an AMD Ryzen™ Threadripper™ PRO CPU, 256 GB ECC RAM and a 2 TB NVMe SSD, connect by remote desktop and work on it like your own workstation, with full administrator access. Your server image is saved between sessions, so your plugins and settings stay installed. Cinema 4D and Redshift are pre-installed on every server.

Multi-GPU servers are billed per minute (1-GPU servers have a one-hour minimum), and sessions of 3 hours or more get 10% off. Our 4-GPU and 8-GPU servers are liquid-cooled; in our own measurements the GPUs hold 60–70°C under sustained render load. Elite support is available 7 days a week by hotline, live chat, Skype, WhatsApp and email.

This quick start shows how to start a server and render Cinema 4D and Redshift on it:

Two Redshift runs we recorded on our servers: 150 frames at 1080p on 4x RTX 5090 at 40.84 seconds per frame (about 1h 42m), and a 301-frame cloth simulation at 1080p on 1x RTX 5090 at 21.3 seconds per frame:

New users get a 100% bonus on the first top-up (minimum $50). See plans on the Cinema 4D and Redshift render farm page or download the iRender GPU app.

Sources: Maxon: Sampling – Production (Redshift), Maxon: Instances (Redshift for Cinema 4D), Maxon: Instance Object (Cinema 4D).

When Redshift frames in Cinema 4D take too long on one card, iRender’s Cinema 4D Redshift render farm lets you work on a dedicated remote server with up to 8x RTX 5090 and render the same scene across multiple GPUs.