Warpbuild 4 vCPU x64 vs Blacksmith 4 vCPU x64
Blacksmith does each build for 1.1–1.4× less; on time, Blacksmith is faster on the Docker build and tied on the Rust and TypeScript builds. Both 4 vCPU x64.
All three builds 1.3× cheaper on Blacksmith: $0.0728 vs $0.0964Blacksmith 1.4× faster: 8m 12s vs 11m 24s
Rust build 1.1× cheaper on Blacksmith: $0.0175 vs $0.0195about the same time: 2m 20s · 1m 59s
TypeScript build 1.4× cheaper on Blacksmith: $0.0089 vs $0.0129about the same time: 1m 29s · 54s
Docker build 1.4× cheaper on Blacksmith: $0.0464 vs $0.0640Blacksmith 1.4× faster: 5m 19s vs 7m 35s
Warpbuild 4 vCPU
jobs ran from WarpBuild (US) and Amazon.com (US)
Blacksmith 4 vCPU
jobs ran from Secured Servers (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 | Warpbuild | Blacksmith | Gap |
|---|---|---|---|
| Cost | |||
| Cost per Rust build | $0.0195 | $0.0175 | 1.1× Blacksmith |
| Cost per TypeScript build | $0.0129 | $0.0089 | 1.4× Blacksmith |
| Cost per Docker build | $0.0640 | $0.0464 | 1.4× Blacksmith |
| Price per minute | $0.0080 | $0.0080 | ≈ even |
| Real builds, wall time | |||
| Rust build (spotify-player) | 2m 20s | 1m 59s | ≈ even |
| TypeScript build (opencode) | 1m 29s | 54s | ≈ even |
| Docker build (PostHog) | 7m 35s | 5m 19s | 1.4× Blacksmith |
| CPU & memory | |||
| PassMark single-thread (points) | 4,318 | 4,236 | ≈ even |
| PassMark CPU Mark, all cores (points) | 11,261 | 15,036 | 1.3× Blacksmith |
| sysbench CPU, 1 thread (events/s) | 6,288 | 6,419 | ≈ even |
| 7-Zip compression (MIPS) | 24,948 | 27,863 | 1.1× Blacksmith |
| Memory read bandwidth | 27.5 GiB/s | 31.5 GiB/s | 1.1× Blacksmith |
| Measurement | Warpbuild | Blacksmith | Gap |
|---|---|---|---|
| Workspace disk (fio) | |||
| Random read, 4k QD32 (IOPS) | 143,694 | 873,112 | 6.1× Blacksmith |
| Random write, 4k QD32 (IOPS) | 40,569 | 251,762 | 6.2× Blacksmith |
| Random read, 4k QD1 (IOPS) | 22,245 | 84,854 | 3.8× Blacksmith |
| QD1 read latency, p99 | 58 µs | 14 µs | 4.1× Blacksmith |
| Sequential read | 7.8 GiB/s | 6.9 GiB/s | 1.1× Warpbuild |
| Sequential write | 5.1 GiB/s | 1.6 GiB/s | 3.1× Warpbuild |
| Sync write, 4k + fdatasync (IOPS) | 1,980 | 18,024 | 9.1× Blacksmith |
| Network, cache & Docker | |||
| Download from GitHub | 151 MiB/s | 107 MiB/s | 1.4× Warpbuild |
| Cache restore, 4 GB (actions/cache) | 50s | 14s | 3.6× Blacksmith |
| Cache save, 4 GB (actions/cache) | 15s | 8s | 1.9× Blacksmith |
| docker pull | 11.9s | 6.5s | 1.8× Blacksmith |
| Database | |||
| pgbench read/write (TPS) | 2,659 | 11,775 | 4.4× Blacksmith |
| pgbench read-only (TPS) | 25,697 | 89,914 | 3.5× Blacksmith |
| Queue | |||
| Queue under burst load | 13s | 11s | 1.2× Blacksmith |
- ≈ even: under 5% apart, or a build inside both runners' run-to-run spread
- Queue: identical jobs queued at the same instant (burst suite)