Hyperion started as a passion project, back in 2016, and is still worked on daily. Our aim with Hyperion is to offer a high fidelity gaming experience even on low-end hardware using our in-house baking system to prepare as much of the lighting and effects as possible ahead of time.
 Glimmer - Emissives |
 Multiplayer, in play-in-editor mode |
- Clustered deferred shading supporting a large number of dynamic lights while maintaining good frame times. Uses forward clustered shading for translucent materials.
- Visual editor on Windows and macOS, built with Avalonia.
- Offline lightmapper integrated into the editor. Bake lightmaps into the scene, reflection/irradiance probes for dynamic objects, fog volumes, and other static lighting data such as shadow maps.
- Real time global illumination and reflections via ray tracing on RT capable hardware, with screen-space options for reflections.
- Glimmer GI - our (non-hardware RT required!) global illumination system. Probe-based diffuse lighting from software ray tracing against a cached scene representation, with a ground clipmap for terrain and incremental relighting as lights and the sun change.
- Rendering backends: DirectX 12 (the default on Windows) and Vulkan (macOS via MoltenVK, Android, and Linux in future). DirectX 12 is a fully supported, mainline backend.
- Shader compiler system with built in permutations support, and live reload in editor to see changes as you make them.
- Scripting via the Strata programming language - JIT compiled, live reload in editor, or AOT linking with shipping builds
- Level streaming via grid-based streaming
- Basic multiplayer setup, with a dedicated server, client-side prediction, replication, etc
Currently, we are focusing our efforts on developing the engine for Windows, macOS, Android, iOS, and Steam Deck via Proton. Editor support is available on Windows and macOS. Linux support is planned.
Documentation is a WIP but we have some available on our site: Documentation For any questions you have, please feel free to ask away in Discord.
AI can a useful tool, but intention is important when building something super complex like a game engine. We aim to keep the usage of AI-generated code limited to bug fixes, UI changes, and occasionally, prototyping / boilerplate. In an effort to maintain transparency, we aim to note where AI is used generally as part of a commit message, comment on the commit on GitHub, or on the PR; although this is on a best effort basis.
We expect contributors to be honest and upfront about the usage of AI generated code and while AI-generated code isn’t de facto unacceptable, the code must pass the same quality bar that human generated code would.