Building physics | Simulation | Energy | Radiation
Procedural Compute is a building simulation task manager.
Procedural Compute orchestrates and scales simulation tasks. It supports a variety of building performance simulation tools.
"We ran a project a month ago on our old computer and it took us 24 hours for one wind direction. With Procedural Compute, the same task can be done on 4 wind directions running parallel in 4-5 hours.
That is a 5 times speedup on one direction, but we get 4 directions out of Procedural Compute, so in total around a 20 times speedup!"
- Mathias Sønderskov Schaltz
Computational Design Lead, Henning Larsen Architects
Procedural Compute is built on top of open source technology.
Procedural Compute are written in Python and C/C++, which can run on any operating system.
Procedural Compute integrates with multiple 3D modelling softwares and allows data transfer between its simulation engines. So all geometry creation, simulation, and visualisation can happen within one interface. It also works and integrates with other open platforms such as Ladybug tools.
Procedural Compute are designed with the purpose of running large scale optimisations on several options at once, giving opportunity to create the multi criteria analysis and wider, deeper analysis of building performance.
Procedural Compute natively supports scaling across multiple machines, making it useful for any number and size of projects.
Application Programming Interface
Procedural Compute is designed as an API first platform, giving you the freedom to integrate it into your company's tools or write your own extensions.
Do you want a private Compute cloud?
Local, in the cloud or hybrid computation?
We know how to build it. We have the experience and know-how when it comes to building and servicing hardware and deploying Procedural Compute onto them.
Do your company need a complete off-grid in-house compute server solution, a fully autonomous online cloud service or anything in-between?
We can help setting it up and maintain it.