Skip to content
Vrynthia ecosystem
Menu

Independent studio · Amsterdam

We build the engine, the pipelines and the worlds that run on them.

Vrynthia Interactive is a game development studio working on its own production technology: Vrynthia Engine v4.2, a distributed rendering infrastructure, and the internal tooling our teams use every day across five countries.

Engine
v4.2
Founded
2017
Render nodes
1,024
Studio site
5 countries

What we do

Technology first, content on top of it.

Most studios rent their pipeline. We built ours because the productions we want to make do not fit inside off-the-shelf tools: open environments with kilometre-scale sightlines, physically plausible light, and teams that are never in the same room at the same time.

That work is split across three teams — Engine, Rendering & Pipeline, and Tools & Telemetry — and it ships as one product: the studio's own engine, with the cloud infrastructure and the companion tooling around it.

Everything we publish here is internal technology. Game builds, raw assets and tools are distributed only to verified partners through our build repository.

Vrynthia Engine v4.2

A production engine for long-sightline worlds

Four years of iteration since v3, rebuilt around streaming and deterministic simulation.

01 Scene submission DCC bundle · 10 GbE 02 Residency compiler pages · light · acoustics 03 Render bridge local or cloud cluster 04 Client viewport HTTP/2 · 60 FPS Deterministic command log — one replayable path from authoring to frame
Engine data path, v4.2. Every stage is the same codebase; the cloud bridge is a transport, not a second renderer.
  1. 01

    Deterministic simulation core

    Fixed-tick world state with reproducible replays. A build played back a week later resolves the same collisions, the same light bakes and the same AI decisions.

  2. 02

    Streaming-first geometry

    Source scenes are compiled into residency pages. The runtime asks for what the camera will need, not for what the level contains, which is what makes kilometre-scale vistas practical.

  3. 03

    Distributed cloud infrastructure

    Long renders, mezzanine encodes and QA distribution run on the same Amsterdam-based cluster, over the same HTTP/2 transport our partners already use for review streams.

At a glance

The platform behind the studio

Same numbers our partners see in the API and on the status page — nothing here is a marketing figure.

Render cluster

1,024

GPU nodes in the shared queue. Mezzanine frames leave the cluster at 85 Mbps per developer stream, over HTTP/2.

Cloud render portal

Telemetry heartbeat

GET /api/v2/telemetry/heartbeat

{
  "ok": true,
  "node": "eu-ams-01",
  "queue_depth": 12,
  "render_nodes": 1024
}

Edge

AMS

One European edge serves three gateways — studio, cloud and builds.

API reference

Companion app

Vrynthia Companion App & Mobile Telemetry

The queue, the builds and the logs — on the phone, not in a terminal.

The companion app for Android and iOS watches the render queue from anywhere: which shot is running, which node it landed on, how long the remaining frames will take. When a build turns green, a push notification arrives before the producer has to ask.

Debug logs are synchronised in the background while the device moves between Wi-Fi and mobile networks. The app picks up where it stopped, so a long capture survives a train ride through four dead zones.

Telemetry is the only channel the app talks to, and it is authenticated with the same developer credentials as the streaming portal.

Journal

What we have been doing

Releases, infrastructure changes and post-mortems — written when there is something to say, not on a schedule.

All entries

Access

Developer and partner accounts are issued, not registered

A Vrynthia System Administrator verifies the studio or partner, creates the account and provisions a hardware token. The request form is open to anyone with a corporate address and an invite code.

Request access Sign in