Jump to ...

Benchmarks

Here are some benchmarks that compare the performance and cost of GitHub-hosted runners and Sprinters-powered AWS runners.

By default, the following configurations are used:

Also note:

• EC2 instances launched in the us-east-1 region.
• EBS volumes are gp3 with 3000 IOPS and 150 MiB/s throughput.
• Spot savings are calculated based on the historical average for this instance type in this region.
• If a configuration differs from the ones above, the different setting sill be shown in the benchmark.

django / django

Benchmark Workflow

github-4vcpu

6m 28s
$0.11200

sprinters-mxl

5m 13s
on-demand $0.02948	1.3x faster and 3.8x cheaper
spot $0.01761	1.3x faster and 6.4x cheaper

sprinters-c4xl

2m 45s
on-demand $0.04408	2.4x faster and 2.5x cheaper
spot $0.02203	2.4x faster and 5.1x cheaper

gohugoio / hugo

Benchmark Workflow

github-4vcpu

43m 56s
$0.70400

sprinters-mxl temp=32

36m 50s
on-demand $0.13134	1.2x faster and 5.4x cheaper
spot $0.05812	1.2x faster and 12.1x cheaper

sprinters-c4xl temp=32

25m 41s
on-demand $0.30622	1.7x faster and 2.3x cheaper
spot $0.11516	1.7x faster and 6.1x cheaper

spring-projects / spring-boot

Benchmark Workflow

github-4vcpu

5m 4s
$0.09600

sprinters-mxl

4m 29s
on-demand $0.02623	1.1x faster and 3.7x cheaper
spot $0.01633	1.1x faster and 5.9x cheaper

sprinters-c4xl

2m 24s
on-demand $0.04408	2.1x faster and 2.2x cheaper
spot $0.02203	2.1x faster and 4.4x cheaper

torvalds / linux

Benchmark Workflow

github-4vcpu

10m 11s
$0.17600

sprinters-mxl

8m 59s
on-demand $0.03922	1.1x faster and 4.5x cheaper
spot $0.02141	1.1x faster and 8.2x cheaper

sprinters-c4xl

3m 12s
on-demand $0.05544	3.2x faster and 3.2x cheaper
spot $0.02605	3.2x faster and 6.8x cheaper