Streaming Learning Center Where Streaming Professionals Learn to Excel
Related Articles
InterDigital’s acquisition of Deep Render marks one of the first instances where a traditional codec IP powerhouse has invested in AI-native compression. Deep Render was built from a clean slate, with an architecture that doesn’t trace its roots to the block-based model that has defined every standard codec for the last 30 years. At some point, that old architecture will be abandoned. Whether Deep Render’s technology is what finally replaces it is too early to tell, but InterDigital is making an interesting bet.
Industry analyst Andrew Garrett offered a useful perspective on LinkedIn (see featured image), noting that Deep Render’s upward trend in the Patent Asset Index was likely a factor in the acquisition. That’s not just a trivia point; it underscores that InterDigital wasn’t chasing hype or a demo, it was buying proven intellectual property strength. Deep Render’s patent portfolio has grown steadily, and in Garrett’s view, its “technology health” made it a strong M&A target. In other words, this deal is as much about owning the IP foundation of AI-based compression as it is about advancing the technology itself.
I was fortunate to serve as a consultant for Deep Render, helping the company highlight its technology and accomplishments. My report on their codec architecture is here (downloaded over 1850 times), along with a demo of their FFmpeg workflow, the company’s own demo, and an interview with CEO Chris Besenbruch. Together, those posts have generated well over twenty thousand impressions across LinkedIn, YouTube, and the Streaming Learning Center, proof of the interest this technology has sparked in the streaming and compression communities.
What Comes Next?
The core question now is what InterDigital plans to do with it. Several paths seem plausible. InterDigital is already a major contributor to the Joint Video Experts Team (JVET) working on the Beyond VVC standard (H.267) across both the Enhanced Compression Model (ECM) and Neural Network Video Coding (NNVC) research tracks. Adrian Pennington covered this in his excellent article, entitled “AI Helps InterDigital Reach Beyond VVC in Race to Develop Next-Gen Codec,” published days before the acquisition was announced.
They could inject some of Deep Render’s AI work into H.267 development, helping shape the next generation of MPEG standards. They could push to make H.267 itself partially AI-native, blending traditional block coding with neural-network-driven prediction and reconstruction. Or they could go fully proprietary, building a parallel AI-based codec family that bypasses the standards process entirely. InterDigital has played all three games before—it holds essential patents in existing standards and also experiments with independent IP tracks.
The appeal of Deep Render’s approach is that it’s not theoretical. It runs in FFmpeg, plays in VLC, and in independent testing has outperformed state-of-the-art encoders like SVT-AV1 and VVenC on key efficiency metrics in very specific use cases. Deep Render showed that AI-based compression can work outside the lab, at real resolutions, with measurable quality and performance gains. That’s rare. Most AI codec experiments still struggle to demonstrate generalizability or practical deployment potential.
Which isn’t to say that the success of Deep Render’s codec is assured. Deep Render made a bet on neural programming-based (NPU) decoding, rather than custom silicon or software. While the codec performs brilliantly on Apple NPUs, it’s unclear whether it will also play on NPUs or similar hardware from other vendors. In markets like smart TVs, where NPUs are not generally deployed, Deep Render won’t play a role for at least 5-7 years, like VVC and AV2. So, despite the early lead and impressive technology, the success of the Deep Render codec isn’t clear.
All that said, while the purchase price wasn’t disclosed, the signal is clear. Deep Render hit enough technical and IP milestones to get InterDigital’s attention, and that’s no small thing. If you’re a technical startup, the lesson is apparent, if hardly a new revelation. Take your IP strategy as seriously as your product strategy. Deep Render did, and it paid off.
InterDigital now has two bets: if Deep Render’s codec takes off, they’re ahead of the curve; if it doesn’t, the patents alone will likely justify the deal, along with access to Deep Render’s excellent technical team. Either way, it’s a smart hedge on where video compression is heading next.
Haha, classic Deep Render! Always proving AI compression isnt just theoretical, even when running on someone elses NPU dream team. Love how InterDigital snags the IP – smart play, like buying the future before it needs its own compression. Cant wait to see if this becomes the new H.267 or just another smart bet on a future tech that might finally run smoothly on NPUs, unlike smart TVs. Good luck Deep Render, hope your codec enjoys the journey!