Subvolts take Learn how Meta cut Instagram video compute time by 94% and scaled the fix. Jan 25, 2023 - Developers
- Learn about optimizing video processing pipelines and the complexities of deploying changes at massive scale.
- Business
- Understand how efficiency gains in core infrastructure can lead to significant cost savings and better user experiences.
This episode details a significant engineering achievement in optimizing resource usage for a major social media platform.
48: A 94% reduction for basic video compute time on Instagram
Audio streams from Meta. Your place is remembered on this device.
From Subvolts
Summary and analysis by Subvolts editorial, AI-assisted.
The Meta Tech Podcast episode "48: A 94% reduction for basic video compute time on Instagram" features Ryan discussing how his team drastically cut video encoding resources for Instagram. The episode explores the technical challenges and the process of rolling out changes that impact billions of users, covering codecs, tradeoffs, and testing.
Imagine making videos load much faster and using less computer power on Instagram. This episode explains how engineers found a clever way to shrink the work needed to prepare videos for everyone to watch, saving lots of energy and time.
This episode offers a deep dive into optimizing video processing for a massive platform like Instagram. It's for engineers and product managers interested in efficiency gains and large-scale deployment. Meta's work here connects to broader industry efforts in efficient AI model deployment and resource management, crucial for scaling AI services.
- HostPascal Harris
- GuestRyan
- Episode Title: 48: A 94% reduction for basic video compute time on Instagram
- Show: Meta Tech Podcast
- Length: 37 minutes
- Compute Reduction: 94%
- Platform: Instagram
Ryan and his team found a quick way of reducing the compute resources spent on encoding videos for Instagram by 94%, but that was actually the easy part. Tune in to learn what the fix was and how you roll out changes that can affect the user experience of billions of users.
Got feedback? Send it to us on Twitter (https://twitter.com/metatechpod), Instagram (https://instagram.com/metatechpod) and don't forget to follow our host @passy (https://twitter.com/passy and https://mastodon.social/@passy). Fancy working with us? Check out https://www.metacareers.com/.
Links:
- Reducing Instagram's basic video compute time by 94 percent - Meta Engineering Blog: https://engineering.fb.com/2022/11/04/video-engineering/instagram-video-processing-encoding-reduction/
- The Diff: https://thediffpodcast.com/
- Unix Signals in Production - Dangers and Pitfalls: https://developers.facebook.com/blog/post/2022/09/27/signals-in-prod-dangers-and-pitfalls/
- Introducing Velox: An open source unified execution engine - https://engineering.fb.com/2022/08/31/open-source/velox/
Timestamps:
- Intro 0:06
- Intro Ryan 1:40
- Transcoding Video at Instagram 2:52
- Codecs and Tradeoffs 5:33
- Client Support 7:13
- Where did the compute go? 9:15
- ABR 10:59
- Progressive/Non-ABR Encodings 12:31
- Saving Encoding Time 13:10
- Testing the Changes 17:39
- Results 26:43
- Popularity Predictions 28:32
- Outro 36:31
38 min41 min87: Everything is Power: Battery Life Engineering for AI Glasses
WithRey, Dmitrii
43 min86: A Hard Cell: Engineering Ultra-Narrow Batteries for AI Glasses
WithKarthik, Myuran
39 min85: Reel Friends: Building Social Discovery that Scales to Billions
WithJoseph, Subasree
Questions people ask
- How was Instagram's video compute time reduced?
- The team found a quick way to reduce compute resources spent on encoding videos for Instagram by 94 percent.
- What are the challenges of rolling out changes for billions of users?
- The episode discusses the process of implementing changes that can affect the user experience of billions of users.
Page generated Oct 19, 2026. Summaries are Subvolts' own; the audio, artwork and notes belong to Meta.