Subsurface models, energy-system studies and fusion research combine large numerical problems with substantial input and checkpoint data. Data preparation and storage access can limit the pace of discovery.
System considerations
Measure read, checkpoint and restart phases separately from compute. Include concurrent users and file counts in storage tests; headline bandwidth alone leaves out metadata and contention.
Evidence to bring
- Dataset size, file count and retention
- Checkpoint and restart window
- End-to-end data and compute timeline
This is an illustrative research or industry context, not an available World Supercomputing service or a promised scientific outcome.
PRIMARY SOURCES · CHECKED 9 NOVEMBER 2026
NERSC: parallelism →Living documentation; checked 2026-11-09 · Strong scaling fixes total problem size; weak scaling fixes work per processor.
NERSC: job best practices →Living documentation; checked 2026-11-09 · Facility-specific operational guidance; use local site policies for actual jobs.