We believe 3D data has a “plumbing” problem. The visuals may be spectacular, but the files behind them are huge. Getting complex scenes from storage to the GPU fast enough has become a real bottleneck—not only for games and film, but also for simulation, robotics and AI.
That is why In Group is backing Zibra AI.
The simplest way to describe what the team is building is “JPEG for 3D.” Zibra compresses complex 3D data so it can be stored, moved and used more efficiently, allowing teams to work with more detail without waiting for infrastructure to catch up.
It is already proving its value in real-time VFX, where fire, smoke and fog can look cinematic without bringing performance to a halt. But the opportunity is much bigger: any industry working with large, fast-moving 3D datasets stands to benefit.
Zibra’s GPU-native compression technology is designed to stream and decompress 3D data at up to 500 GB/s. That means less data moving through the system and more computing power available for the work that matters.
There is a sustainability angle here, too. Smaller files can reduce the storage, network capacity and computing needed to achieve the same result. As 3D and AI workloads continue to grow, making that infrastructure lighter and less energy-intensive truly matters.
We like companies solving the unglamorous but critical problems. The ones that make the next generation of products possible. Zibra AI is one of them.
