Benchmarks¶
Altius L1 includes a built-in benchmarking tool under Performance → Benchmarks in the operator dashboard, so you can measure throughput and latency against a running node from a point-and-click interface — nothing to build or run yourself. Results are tracked over time and can be exported to share.
Figures are targets, not published results
The numbers below are targets / representative benchmarks. Run your own benchmark on your deployment and hardware before citing any specific figure externally.
Targets¶
| Metric | Target |
|---|---|
| Block production | 60 ms |
| Sustained throughput | 4,000+ TPS |
| Gas throughput | 1.6M+ gas/s |
| Confirmation latency | < 1 ms |
| Execution | Parallelized EVM (Prague) |
| Fees | Multi-token & sponsored fee support |
Deployments can be tuned to your use case, or hardware-optimized to exceed these numbers.
What we measure¶
- Throughput (TPS) — submitted vs. confirmed transactions per second under load.
- Confirmation latency — tracked at percentiles (p95 / p99) rather than averages.
- Gas throughput — gas processed per second.
- Block production cadence — configurable on a sub-second basis for predictable, low-latency finalization.
Benchmark from the dashboard¶
The dashboard's Performance → Benchmarks view runs benchmarks against your own network and shows the results live — you don't need access to the Altius codebase.
- Open the Altius dashboard and go to Performance → Benchmarks.
- Choose a load profile — Smoke, Medium, or Stress — or define a custom run (target TPS, duration, sender count, in-flight limit). A Max-TPS mode probes peak sustained throughput.
- Start the run and watch it live: submitted and confirmed TPS against your target, plus confirmation latency at p95 / p99.
- Compare each run against your SLO targets, browse the History of past runs, and export results as JSON or Markdown to share with your team.
Note
Running benchmarks uses the dashboard's administration controls, so you'll need dashboard access with the appropriate role. The available profiles and SLO thresholds depend on your deployment's configuration.