Talik is highly optimized. Whether you are recycling a decade-old desktop for a small shop or spinning up a massive 64-channel perimeter on modern silicon, Talik scales with you.
Talik is currently deployed exclusively as a Docker container, maximizing efficiency and isolating dependencies. It requires a Linux host environment.
For bare-metal deployments, we strongly recommend CachyOS. Its highly tuned kernel, BORE scheduler, and optimized packages provide measurable performance gains for continuous AI inference and video writing.
* Note: A native Windows deployment is actively in development and coming soon.
NVR workloads are intensely write-heavy. Standard desktop hard drives will fail prematurely.
Talik's containerized architecture allows it to run on hardware that other modern VMS platforms have abandoned. Perfect for small footprint deployments.
Scaling up the camera count requires slightly more memory for pre-roll buffering and a modern quad-core to comfortably handle concurrent WebRTC streams and AI tasks.
For medium-to-large facilities. An octa-core processor ensures that multiple users can view dynamic grids and scrub the timeline simultaneously without impacting the autonomous recording engine.
The ultimate deployment. Massive concurrent write I/O and intensive parallel AI inference. High-thread-count modern silicon is required to maintain real-time analysis across dozens of high-definition feeds.