AI & big data

Datasets and checkpoints between regions, clouds and clusters

Stop paying for idle GPUs while data copies. fft moves training data and checkpoints between regions, clouds and on-premises clusters at up to 94 Gbps, from the machines you already run.

$ fft send ./checkpoint-step-120000 fft:eyJ2Ijox… -r

100 GiB in 11 sEnd to end, encrypted and verified

91–94 GbpsSustained on 100 Gbps networking, encrypted

≤310 MBPeak memory per side, whatever the file size

Fills 10–100 Gbps pipes

Uses the bandwidth ordinary tools leave idle, with nothing to tune.

Light on the machine

Flat memory and hardware-accelerated crypto leave room for the work.

Script- and agent-ready

Machine-readable everything, and the full CLI contract as JSON.

Picks up where it left off

An interrupted transfer resumes; nothing delivered is resent.

Related use cases

Research & education

Instrument data and datasets across 10–100 Gbps research networks.

Backup & replication

Off-site and cross-region copies that finish inside the window.

Biomedical & life sciences

Sequencing runs, microscopy and imaging data between instruments, cores and clusters.

Your machine

End-to-end encrypted

Their machine

Falcon fft and Makai Systems: not in the path

No middleman. We never see your data.

Every transfer goes peer to peer, straight from one machine to the other, encrypted with keys that exist only on those two machines. There is no Falcon cloud, relay or storage in between, so neither Falcon fft nor Makai Systems ever sees your files, their names, or who you send them to.

Try it on your own network

Falcon fft is in early access. Leave your email and we'll set up a trial on your own links.