Results with the test beside them.
Engineers, researchers, and enterprise buyers
These published summaries report individual Rubian and Kestowv measurements. Verification requires the original run record, tested release, workload definition and correctness output. The methodology explains how to evaluate those records.
69.669x faster directory listing.
Median result for one in-process Rubian operation against the named process-start comparison. The result doesn’t support a generic shell or operating-system superiority claim.
5,082.148 bridge requests per second.
One qualification completed every accepted request with zero correctness errors.
1,512,206 work units per second.
One equal-work Kestowv control completed every accepted work unit with matching checksums. The result belongs to that named control condition.
Read the result in context.
Use each summary to identify the measurement of interest. Obtain its original run record before treating the figure as reproduced evidence or comparing it with another system.
Project Valux needs its own results.
Existing Rubian and Kestowv measurements don’t automatically describe Project Valux. New claims require matched work, correctness checks, telemetry, and repeated trials.
Working through a hardware or runtime decision? Explore compute evaluations
Startup and discovery, file and project organization, Forge libraries, Taskman, agent work and Kestowv pilot operations.