Notes on moving data at scale.
Engineering and field notes on high-speed, petascale data movement — from the team that set the record.
Articles and guides
PACS and VNA migration at scale.
Why imaging-archive migrations take months, how to separate the bulk copy from DICOM reconciliation, how long a petabyte takes, and a cutover checklist.
Read → Guide · Nov 18, 2026How to keep GPUs fed.
Where GPU time leaks to data movement — staging, checkpointing, and moving data between sites and clouds — what it costs, and a playbook.
Read → Guide · Nov 18, 2026Managed file transfer vs. high-speed data movement.
Two tools that both move files but do very different jobs: what each is built for, the signs you have outgrown MFT for bulk data, and how to run both.
Read → Guide · Nov 18, 2026Ship drives or use the network?
When it is faster to put data on a truck, when the network wins, how much a link moves per day at line rate, and a five-minute decision checklist.
Read → Guide · Nov 18, 2026How to transfer petabytes of data.
How long a petabyte really takes, why big transfers run slower than your link, the options compared, and a checklist for reaching line rate.
Read → Comparison · Mar 31, 2026rsync, Globus, or an appliance?
The tools that move a few gigabytes don't move a few petabytes. How rsync, Globus, Aspera, DIY, and an appliance stack up when the data gets big.
Read → Architecture · Jan 20, 2026Burst buffers, explained.
A 100 Gbps link is only as fast as the storage feeding it. How an NVMe burst buffer keeps your network running at line rate instead of stalling on disk.
Read → Engineering · Oct 14, 2025The network isn't the bottleneck — the whole path is.
Buy a faster link and your data moves faster, right? At petabyte scale, usually not. What actually limits end-to-end data movement — and why the fix is co-design.
Read →