TL;DR
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
A benchmark by Tony Metzidis comparing WSL 3.0.2 with WSL 2.6.3.0 found performance gains ranging from about 4% in basic system-call throughput to 61% in a memory-bandwidth test. A GoReleaser build took about 4% less wall-clock time, showing that results depend on the workload and test setup.
WSL 3.0.2 outperformed WSL 2.6.3.0 in several tests conducted by developer Tony Metzidis, with reported differences ranging from about 4% in a basic syscall benchmark to about 61% in memory bandwidth. A GoReleaser compilation finished about 4% sooner by wall-clock time, so the results suggest the upgrade’s benefit depends on the workload rather than delivering one uniform speed increase.
Metzidis compared WSL 2.6.3.0, running Linux kernel 6.6.87.2, with WSL 3.0.2.0, running kernel 6.18.40.1. The supplied report describes a test on a Windows 11 Pro desktop with an Intel Core i5-8500T, 16 GB of RAM, and WSL limited to two processors and 4 GB of memory. Both test environments used Alpine Linux 3.23.0.
In the reported low-level tests, memory-copy bandwidth rose from 7.84 GB/s to 12.64 GB/s, a 61.35% increase. A process ping-pong test rose 12.68%, while context-switch latency fell from 20.64 to 18.32 microseconds per operation. Hackbench, a messaging test, took about 10% less time. By comparison, getppid syscall throughput increased 3.98%.
The real-world build test showed a smaller overall change. Building GoReleaser took 183.80 seconds on WSL 3 versus 191.27 seconds on WSL 2, a reduction of 7.47 seconds, or 3.91%. The report also records a 15.98% reduction in system time, the time the processor spent executing kernel code. These figures describe one test environment and workload, not a guaranteed result for every WSL installation.
Where WSL 3 Gains Matter Most
The test points to a meaningful distinction for developers: memory-heavy and task-switching workloads may benefit more than jobs dominated by raw computation. Applications that move data frequently or coordinate many processes could see improvements in relevant operations, while the overall speedup for a compiler build may remain modest.
That difference matters when evaluating an upgrade. The largest benchmark result is not the same as a 61% reduction in application runtime: it measures memory-copy bandwidth in a specific microbenchmark. In the GoReleaser test, total elapsed time improved by about 4%, despite a larger reported reduction in kernel time. Readers should treat the figures as measurements from one machine, not a general performance guarantee.
As an affiliate, we earn on qualifying purchases.
The Kernels and Test Setup
Metzidis’s report says WSL 3.x uses a newer virtualization runtime and advances the shipped guest kernel from the 6.6 LTS branch to kernel 6.18. The comparison therefore measures the tested WSL releases and their associated software stack together; it does not isolate the effect of a single kernel change from other runtime or configuration differences.
The machine used an Intel Core i5-8500T, with six physical cores, but the WSL configuration was capped at two processors and 4 GB of memory. The report says the system ran Windows 11 Pro with virtualization-based security active. For the Go build, the compiler cache was cleared while the module cache was primed. Those details help describe how the test was run, but results could differ with other processors, resource limits, operating-system builds, projects, or benchmark methods.
“where you see the benefit depends heavily on your workload profile”
— Tony Metzidis, in the benchmark report
As an affiliate, we earn on qualifying purchases.
Limits of the Reported Comparison
The supplied material contains results from one host computer and one set of resource limits. It does not establish whether the same gains appear across a broader range of processors, Windows versions, WSL configurations, or common developer applications. The report also does not provide independent replication or a wider set of end-to-end workloads.
The benchmark author links some measured changes to kernel and virtualization behavior, including page reporting and interrupt handling. Those are explanations offered in the report; the figures alone do not establish how much each individual change contributed. It is also unclear from the supplied material whether the test has been independently reviewed or reproduced.
As an affiliate, we earn on qualifying purchases.
More Workloads Could Clarify the Gains
The next useful step would be repeatable comparisons on additional hardware, with more end-to-end workloads and clearly documented configurations. Tests of different build systems, storage-heavy projects, containers, and memory-intensive applications could show whether the microbenchmark gains translate into routine development tasks.
No broader follow-up or independent replication is identified in the supplied report. Until more comparisons are available, users can treat these measurements as an indication of how WSL 3 performed on Metzidis’s setup, then benchmark their own workloads before drawing conclusions about expected time savings.
Windows Subsystem for Linux development
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
How much faster was WSL 3 in the report?
The reported changes varied by test: memory-copy bandwidth increased 61.35%, while getppid syscall throughput increased 3.98%. The GoReleaser build’s wall-clock time fell by 3.91%.
Did the GoReleaser build finish 60% faster?
No. It finished in 183.80 seconds on WSL 3 compared with 191.27 seconds on WSL 2, a reduction of 7.47 seconds, or 3.91%.
What versions were compared?
The report compared WSL 2.6.3.0 with Linux kernel 6.6.87.2 against WSL 3.0.2.0 with Linux kernel 6.18.40.1.
Will every WSL user see the same performance gains?
The report does not establish that. It tested one desktop, a two-processor WSL allocation, and a limited set of benchmarks, so results may differ with other hardware, settings, and workloads.
Source: hn
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
