A controlled Xcode Derived Data cleanup experiment
In one disposable Xcode sample, clearing its dedicated DerivedData folder removed generated output and required a rebuild. The unchanged build was faster than the build after cleanup in this run. This illustrates a tradeoff; it does not predict your project’s build time, typical storage savings or CleanMacster’s cleaning speed.
Environment and fixture
Mac16,9, Apple M4 Max; macOS 27.0.1 (26A434); Xcode 27.2, Build version 27B48. Observed October 25, 2026, Europe/Istanbul.
One generated macOS C command-line target, Debug configuration, no packages, signing disabled. This is a small controlled fixture, not CleanMacster or a real customer project. No real archives, simulator data or project output were removed.
Procedure
- Create the sample project using the specification in the evidence.
- Build with its own
-derivedDataPath ./DerivedData. - Build again unchanged and time the command with a monotonic clock.
- Measure the sample’s DerivedData using
du -sk. - Remove only that temporary fixture directory.
- Build the same target again.
xcodebuild -project StorageFixture.xcodeproj -scheme StorageFixture -configuration Debug -derivedDataPath ./DerivedData CODE_SIGNING_ALLOWED=NO buildObserved time
| Stage | Elapsed wall time |
|---|---|
| First build | 2.903 s |
| Unchanged build | 0.855 s |
| Build after removing fixture Derived Data | 1.897 s |
Storage observations
- Selected and removed allocated fixture output: 19,243,008 bytes.
- Allocated fixture output after rebuilding: 19,238,912 bytes.
- System-reported free-space delta immediately after removal: 19,275,776 bytes.
The last number includes unrelated background activity and is not a causal estimate of recovered capacity. Selected allocated bytes and observed system free-space change are different measurements.
Limitations
One run per stage provides no statistical confidence. Compiler and operating-system caches outside Derived Data remain warm. A tiny C target does not model a large Swift workspace or indexing costs. We did not test a CleanMacster cleanup operation or compare competing products.
Reproduce and inspect
Download the fixture specification, source and aggregate evidence. Use a new temporary directory and keep output separate from real projects. The full procedure is retained in the repository’s research script.
Sources and verification
Sources checked . CleanMacster publishes this page; see our editorial policy. Instructions can differ by system and app version.
Review your Mac before cleaning
Scanning is free. Cleaning requires a license. For macOS 14 or later.