Command Palette

Search for a command to run...

GitHubRun the benchmark
Providers / Vercel

Vercel

Frontend cloud; measured here via Sandbox, its ephemeral microVM compute.

Compute

Sandbox2 vCPU / 4 GBiad1Compare in Compute
CPU
Intel(R) Xeon(R) Processor @ 2.90GHz
Cores
2
Memory
4 GiB
OS / arch
linux/amd64
Kernel
6.12.76
CLI
0.1.0
runtime
node24
invoked_via
@vercel/sandbox SDK
MetricValuevs best
single_core_hash1,325 MB/sbest
multi_core_hash2,647 MB/sbest
scaling_efficiency99.9 %best

SHA-256 over 64 KiB buffers; uses CPU crypto extensions where available

MetricValuevs best
seq_write1,390 MB/sbest
seq_read3,464 MB/sbest
rand_read_4k73.8K IOPSbest
fsync_latency_p500.03 msbest

reads bypass the page cache (direct I/O) · test file: 1024 MiB, random reads at queue depth 4

MetricValuevs best
copy_bandwidth6.64 GB/sbest
random_access_latency110 nsbest

bandwidth counts bytes copied (read+write traffic is ~2x)

MetricValuevs best
latency_p5027.9 msbest
download3,093 Mbpsbest
upload845 Mbpsbest

measured against the nearest Cloudflare edge

MetricValuevs best
unit_time_p5014.6 msbest
consistency_cv10.5 %best
p99_over_p501.36 ratiobest
cpu_steal0.30 %best

consistency_cv near 0% and p99_over_p50 near 1.0 indicate a quiet host; high values suggest noisy neighbors or throttling

Raw report: data/results/vercel-sandbox-2vcpu.json