Fastest GitHub Actions runners: disk I/O.

How this shows up in real builds →

What the numbers say

Extremes on x64, and RAM against disk.

• On x64, one small read at a time (4k, QD1) comes back 84,602 times a second on Blacksmith 4 vCPU and 1,571 on RunsOn m8i.2xlarge · EBS gp3, 54× apart; p99 latency is 14 µs against 1,036 µs.
• Synced 4k writes on x64 run from 498 per second on RunsOn m8i.2xlarge · EBS gp3 (workspace mounted nobarrier) to 20,425 on Blacksmith 4 vCPU (workspace mounted nobarrier).
• r8a.xlarge with tmpfs ($0.0020/min at spot) answers 1,109,976 QD1 reads and 430,233 synced writes per second: 91× and 36× the best local NVMe at 4 vCPU (i7i.xlarge, $0.0019) and 607× and 797× default EBS gp3 (m8a.xlarge, $0.0015); its workspace is 46 GiB, from RAM. (EC2 storage suite) What that does to real builds.

More findings21: EBS, arm64, NVMe vs EBS, provisioned gp3, tmpfs, storage classes

• Every EBS-backed workspace (3 AWS CodeBuild x64 runners, 25 RunsOn configurations and 3 Warpbuild configurations) answers between 1,560 and 1,844 QD1 reads and 498 to 563 synced writes per second, whoever runs it and whatever throughput it is provisioned for: each request is a round trip to network storage.
• On arm64, one small read at a time (4k, QD1) comes back 26,620 times a second on Warpbuild xfast (Apple M4 Pro) 6 vCPU and 1,560 on RunsOn m9g.xlarge · EBS gp3, 17× apart; p99 latency is 70 µs against 938 µs.
• Synced 4k writes on arm64 run from 537 per second on RunsOn c8g.large · EBS gp3 (workspace mounted nobarrier) to 17,418 on RunsOn m9gd.2xlarge · local NVMe.
• Same Xeon 6975P-C CPU, different disk: m8id.2xlarge on local NVMe answers 10,957 QD1 reads and 7,657 synced writes per second, m8i.2xlarge on EBS gp3 1,571 and 498.
• Same Xeon 6975P-C CPU, different disk: m8id.large on local NVMe answers 11,021 QD1 reads and 5,565 synced writes per second, m8i-flex.large on EBS gp3 1,745 and 532.
• Same Xeon 6975P-C CPU, different disk: m8id.xlarge on local NVMe answers 11,064 QD1 reads and 7,974 synced writes per second, m8i-flex.xlarge on EBS gp3 1,707 and 531.
• Same Neoverse-V3 CPU, different disk: m9gd.2xlarge on local NVMe answers 13,464 QD1 reads and 17,418 synced writes per second, m9g.2xlarge on EBS gp3 1,788 and 542.
• Same Neoverse-V3 CPU, different disk: m9gd.xlarge on local NVMe answers 12,823 QD1 reads and 13,696 synced writes per second, m9g.xlarge on EBS gp3 1,560 and 543.
• Same Neoverse-V3 CPU, different disk: m9gd.large on local NVMe answers 13,492 QD1 reads and 7,267 synced writes per second, m9g.large on EBS gp3 1,788 and 542.
• Provisioning gp3 at 1000 MiB/s on m8a.2xlarge takes sequential reads from 402 MiB/s to 1,002 MiB/s (2.5×); QD1 reads barely move (1,791 → 1,792 IOPS): it buys bandwidth, not latency.
• Provisioning gp3 at 1000 MiB/s on m8azn.xlarge takes sequential reads from 401 MiB/s to 1,001 MiB/s (2.5×); QD1 reads barely move (1,815 → 1,764 IOPS): it buys bandwidth, not latency.
• tmpfs on m8a.2xlarge is not a disk at all: 1,156,332 QD1 reads and 428,514 synced writes per second, from RAM. The workspace is 46 GiB, sized by memory.
• tmpfs on m9g.2xlarge is not a disk at all: 664,023 QD1 reads and 301,438 synced writes per second, from RAM. The workspace is 46 GiB, sized by memory.
• tmpfs on m9g.xlarge is not a disk at all: 680,179 QD1 reads and 303,641 synced writes per second, from RAM. The workspace is 15 GiB, sized by memory.
• tmpfs on r8a.2xlarge is not a disk at all: 1,156,647 QD1 reads and 430,802 synced writes per second, from RAM. The workspace is 92 GiB, sized by memory.
• tmpfs on r9g.xlarge is not a disk at all: 658,442 QD1 reads and 300,910 synced writes per second, from RAM. The workspace is 46 GiB, sized by memory.
• tmpfs: 658k–1.2M QD1 reads and 301k–431k synced writes per second across 6 runners (6 RunsOn configurations).
• VM disk: 6.0k–85k QD1 reads and 578–20k synced writes per second across 42 runners from 7 providers (6 Avrea runners, 6 Blacksmith runners, 3 GitHub runners, 11 Namespace runners, 2 StarSling x64 runners, 9 Ubicloud runners and 5 Warpbuild runners).
• local NVMe: 11k–13k QD1 reads and 5.6k–17k synced writes per second across 9 runners (9 RunsOn configurations).
• network disk: 6.0k–8.9k QD1 reads and 3.1k–4.6k synced writes per second across 2 runners (2 GitHub runners).
• EBS: 1.6k–1.8k QD1 reads and 498–563 synced writes per second across 31 runners from 3 providers (3 AWS CodeBuild x64 runners, 25 RunsOn configurations and 3 Warpbuild configurations).

Reading these numbers

fio measures what a disk can do. Whether a job gets there depends on the job.

At 4 vCPU, Blacksmith answers 46× the QD1 reads of RunsOn m8a.xlarge · EBS gp3 (84,602 against 1,828 per second), and the Rust build takes 1m 59s on one and 1m 55s on the other. The Docker build is where the disk shows: 5m 19s against 9m 05s. Every build, per runner.

What each column says about a job

• QD1 reads and synced writes: databases, SQLite-backed tests and anything that flushes each write. Every flush waits for the storage, and a network disk pays a round trip each time.
• Random writes at depth and sequential throughput: docker build layers, package installs and cache archives. They batch their writes and run into the disk's ceiling, not its latency.
• None of them: compile jobs. A compiler waits for the CPU, and it reads its files back from the page cache.

Every runner

Each provider's best disk first; its other machines sit behind +N more. Bars are log scale.

Pick the disk per job on RunsOn

• The instance type sets the storage under the workspace: EBS gp3 by default, provisioned gp3 for bandwidth, local NVMe on the d and i families, or tmpfs when the job fits in RAM. Pick it with the family label; each one, measured: EC2 storage benchmark.
• What tmpfs does to real builds, what it costs and where it does nothing: Disks? Where we're going, we don't need disks.

How it's measured

Methodology →

fio runs inside the job, in GITHUB_WORKSPACE: the disk a build writes to, on the filesystem the provider mounted there. Every runner is compared, whatever its shape.

fio settingsblock sizes, queue depths, mounts

• QD1 random read: 4k blocks, one request in flight; IOPS and p99 latency.
• Sync write: 4k writes, each followed by fdatasync.
• Random read and write: 4k blocks, 4 jobs with 32 requests in flight each.
• Sequential read and write: 1M blocks, 32 requests in flight.
• The workspace filesystem, its size and its mount options come from the job's own mount table.
• Earlier versions of this page used a different harness and other fio settings on 2 vCPU runners; those numbers are gone rather than mixed in.