VSNexus Technologies

Benchmark

Clean-channel encode and decode, 20 KB to 1 GiB

Verified (software)

Clean-channel encode and decode, 20 KB to 1 GiB20 KB encode: 0.06; 20 KB decode: 0.08; 1 MiB encode: 0.21; 1 MiB decode: 0.44; 10 MiB encode: 1.25; 10 MiB decode: 2.06; 100 MiB encode: 10.9; 100 MiB decode: 18.4; 1 GiB encode: 105.8; 1 GiB decode: 195.4Seconds (log)0.060.0820 KB0.210.441 MiB1.252.0610 MiB10.918.4100 MiB1061951 GiBInput size
Verified (software)encodedecodeMethodology and limitations
Result
1 GiB: encode 105.8 s, decode 195.4 s, byte-exact; 9.78 nt per byte (seconds)
Classification
Checked by a committed test or script in the repository. Software behaviour only.
Methodology
Wall time and peak RSS per stage, measured in a fresh process per stage; exactness checked by byte comparison of the restored output.
Dataset
Random input files of 20 KB, 1 MiB, 10 MiB, 100 MiB and 1 GiB.
Configuration
V9 pipeline, profile v4-balanced, 4 workers; each stage in its own process. Every strand read once with no errors.
Environment
Development host: shared VPS, Intel Xeon Gold 6240 @ 2.60 GHz, 4 cores / 8 threads, Ubuntu 24.04, Python 3.12, CPU only. Host load 2.2 at the start of the run and 1.7 at the end.
Limitations
A clean channel has no sequencing errors, so the clustering and consensus stages do not run. This measures the codec and I/O path, not recovery from noisy reads. One host, one run per size.
Version
V9
Source
v9.0.0/experiments/v9/scale/results/scale.json
Reproduce
experiments/v9/reproduce.sh scale · how to reproduce