SiMa.ai.
Twenty plus films
for the edge.
A library of short explainer films covering what SiMa.ai’s silicon actually does once it is out in the world, plus a logo animation to open every one of them.
The client.
SiMa.ai build the chips that let machine learning run on the device instead of in a data centre. Their MLSoC and ONE Platform put computer vision and generative models straight into robots, drones, cars, factory lines, shop floors and hospital equipment.
Founded by Krishna Rangasayee and backed by Dell Technologies Capital, Amplify Partners and Fidelity, they sit in one of the harder corners of tech to explain. Silicon that nobody sees, doing work nobody watches.
The problem.
A chip company cannot show you the product. The product is a square of silicon. Everything interesting about it only appears in what it lets somebody else build.
So the job was not one hero film. It was a library. One film per use case, each one short enough to drop into a sales call, a landing page or a feed, each one making a single idea land without a technical background.
You cannot film a chip. You film what it makes possible.
The approach in one lineMade early.
This was made in the first wave of generative video, when the tools had only just arrived and almost nobody was putting them into client work. We were, and we built a whole library with them.
You can see the year in the fidelity. Those tools were rougher then than they are now. What holds up is the structure. Every film still gets one idea across in under a minute, which was the actual brief.
We would make these sharper today. We would not make them differently.
The opener.
How it was made.
Scripted, designed, animated, edited and mixed by us, in collaboration with Altima Creative.
Twenty plus films is a production problem before it is a creative one. The system came first: one grid, one type treatment, one motion language, one opener. After that the films could be made in parallel without drifting apart.