Microsoft’s WSL 3.x release updates the default virtualization runtime and advances the guest Linux kernel from the 6.6 LTS series to version 6.18. To isolate the effect, Tony Metzidis ran side‑by‑side benchmarks using an Alpine Linux 3.23.0 root filesystem as the guest OS, with Go 1.27.1 installed, on an HP ProDesk 400 G4 Desktop Mini (Intel Core i5‑8500T, 6C/6T @ 2.10 GHz base, 16 GB DDR4) under Windows 11 Pro with Virtualization‑Based Security enabled, limiting WSL to two pinned vCPUs and 4 GB of memory via .wslconfig.
Microbenchmark measurements show WSL3 delivering up to a 61.35 % increase in sequential memory bandwidth, rising from 7.84 GB/s with WSL2 to 12.64 GB/s with WSL3, which the author attributes to the page_reporting.page_reporting_order=5 setting that reduces hypervisor trap frequency during dynamic memory reclamation. Context‑switch latency fell from 20.64 µs to 18.32 µs (‑11.25 %), scheduler‑pipe ping‑pong throughput rose from 48,445 to 54,589 ops/s (+12.68 %), and the Hackbench messaging test (20 groups, 800 tasks) completed in 13.05 s versus 14.51 s (‑10.06 %). Basic getppid() system‑call throughput improved by about 4 %.
In a real‑world workload, compiling the GoReleaser codebase with an empty compiler cache and a primed module cache, wall‑clock time dropped from 191.27 s to 183.80 s (‑3.91 %), user CPU time from 298.99 s to 289.38 s (‑3.21 %), and kernel time from 57.39 s to 48.22 s (‑15.98 %). Because the compilation is dominated by user‑space lexing, AST construction, type checking and SSA optimization—over 80 % of cycles—the overall speed is capped by the two‑core CPU frequency, so compute‑bound tasks see mainly reduced kernel overhead while memory‑intensive or IPC‑heavy workloads benefit most from the bandwidth and latency gains.