Blacksmith 4 vCPU x64 vs Namespace 4x16 x64
Blacksmith does the three builds for 1.2× less in total (less on the Rust and Docker builds and more on the TypeScript build); neither is clearly faster (within 5% or inside each other's run-to-run range). Both 4 vCPU x64.
All three builds 1.2× cheaper on Blacksmith: $0.0741 vs $0.0900Namespace 1.2× faster: 7m 02s vs 8m 27s
Rust build 1.3× cheaper on Blacksmith: $0.0155 vs $0.0198within run-to-run spread: 1m 47s · 1m 33s
TypeScript build 1.1× cheaper on Namespace: $0.0106 vs $0.0121within run-to-run spread: 1m 19s · 45s
Docker build 1.3× cheaper on Blacksmith: $0.0465 vs $0.0596within run-to-run spread: 5m 22s · 4m 44s
Blacksmith 4 vCPU
jobs ran from Secured Servers (US)
Namespace 4x16
jobs ran from Namespace Labs (US)
Side by side
Every measurement both runners share, cost first. The dot marks the better value; gap is the larger over the smaller.
| Measurement | Blacksmith | Namespace | Gap |
|---|---|---|---|
| Cost | |||
| Cost per Rust build | $0.0155 | $0.0198 | 1.3× Blacksmith |
| Cost per TypeScript build | $0.0121 | $0.0106 | 1.1× Namespace |
| Cost per Docker build | $0.0465 | $0.0596 | 1.3× Blacksmith |
| Price per minute | $0.0080 | $0.0120 overage rate | 1.5× Blacksmith |
| Real builds, wall time | |||
| Rust build (spotify-player) | 1m 47s | 1m 33s | ≈ even |
| TypeScript build (opencode) | 1m 19s | 45s | ≈ even |
| Docker build (PostHog) | 5m 22s | 4m 44s | ≈ even |
| CPU & memory | |||
| PassMark single-thread (points) | 4,332 | 4,628 | 1.1× Namespace |
| PassMark CPU Mark, all cores (points) | 14,771 | 15,965 | 1.1× Namespace |
| sysbench CPU, 1 thread (events/s) | 6,493 | 6,887 | 1.1× Namespace |
| 7-Zip compression (MIPS) | 28,802 | 33,572 | 1.2× Namespace |
| Memory read bandwidth | 41.7 GiB/s | 44.7 GiB/s | 1.1× Namespace |
| Measurement | Blacksmith | Namespace | Gap |
|---|---|---|---|
| Workspace disk (fio) | |||
| Random read, 4k QD32 (IOPS) | 728,035 | 1,202,685 | 1.7× Namespace |
| Random write, 4k QD32 (IOPS) | 218,915 | 564,373 | 2.6× Namespace |
| Random read, 4k QD1 (IOPS) | 71,728 | 104,443 | 1.5× Namespace |
| QD1 read latency, p99 | 16 µs | 11 µs | 1.5× Namespace |
| Sequential read | 7.0 GiB/s | 5.0 GiB/s | 1.4× Blacksmith |
| Sequential write | 1.7 GiB/s | 3.4 GiB/s | 2.0× Namespace |
| Sync write, 4k + fdatasync (IOPS) | 17,925 | 28,561 | 1.6× Namespace |
| Network, cache & Docker | |||
| Download from GitHub | 76 MiB/s | 187 MiB/s | 2.5× Namespace |
| Cache restore, 4 GB (actions/cache) | 11s | 21s | 1.9× Blacksmith |
| Cache save, 4 GB (actions/cache) | 7s | 13s | 2.0× Blacksmith |
| docker pull | 7.9s | 3.2s | 2.4× Namespace |
| Database | |||
| pgbench read/write (TPS) | 12,408 | 13,662 | 1.1× Namespace |
| pgbench read-only (TPS) | 109,979 | 114,631 | ≈ even |
| Queue | |||
| Queue under burst load | 10s | 10s | ≈ even |
- ≈ even: under 5% apart, or a build inside both runners' run-to-run spread
- Queue: identical jobs queued at the same instant (burst suite)