I recently interviewed Chri Besenbruch, CEO of Deep Render, to discuss their end-to-end AI codec technology. Unlike traditional codecs like HEVC, VVC, and VP9, Deep Render’s AI codec shifts the entire encoding/decoding process to neural networks, promising faster performance and improved efficiency. The YouTube video of our talk is embedded below.
Key Takeaways:
• Fast Encoding & Decoding: Comparable to traditional codecs, with real-time decoding even on devices like the iPhone 13. (Performance metrics at 10:18)
• Quality Comparisons: Claims of 40–50% better efficiency than AV1, with strong performance against HEVC (x265) and VP9. (Details at 07:00)
• AI Codec vs. AI Codec: How Deep Render stacks up against other AI-driven solutions from companies like Microsoft and Qualcomm. (Insights at 14:00)
Chri predicts commercial deployments by 12/2026.
If you’re curious about the future of video compression, this conversation is worth a watch.
Disclosure: Deep Render is a consulting client, and I haven’t independently tested the codec, so I can’t verify these claims.