Your instruments don't wait. Neither should your pipeline. zx Streaming append-streams live data as it is written — no waiting for a complete file — ingesting in parallel at line rate — as fast as your instruments and network allow, scaling out with no software ceiling.
zx Streaming is part of the unified Zettar zx data mover. It uses the same engine behind zx File, zx Object, and zx Single-Site. Instead of moving data after a file closes, it streams every byte as it is produced — built for high-speed sensors, instruments, and genome sequencers. Its design is motivated by the data flood at SLAC LCLS-II.
Append-streams live output as it is written. No waiting for a complete file — transfer starts on the first byte, so nothing sits idle on disk.
Ingests in parallel at line rate — as fast as your instruments and network allow, scaling out with no software ceiling. Fast enough to keep pace with the fastest instruments.
Purpose-built for high-speed sensors, scientific instruments, and genome sequencers that produce data without pause.
Delivers live data to analysis and pipelines in real time, with checksums and TLS in flight — so compute starts sooner and never waits on the wire.
"84 Gbps streaming over a 100 Gbps link, Switzerland to Northern California."
Moving live, fast-growing data as it is written — instead of waiting for a complete file — by ingesting in parallel at line rate. Ideal for high-speed instruments and sensors.
Line rate — it ingests as fast as your instruments and network allow, and scales out with no software ceiling. The 1 PB / 29-hour record proves the engine at petascale.
Sensors, scientific instruments, and sequencers — any source producing live, fast-growing output that has to move the moment it is created.
It keeps pace with the instrument in real time, so downstream processing starts immediately instead of waiting for the run to finish and the file to close.
Yes — unconditional end-to-end checksums and TLS on every transfer, the same integrity guarantees as bulk movement.
See zx Streaming ingest live sensor, instrument, and sequencer output in parallel at line rate — as fast as your instruments and network allow — and feed it downstream in real time.