🚀 The Ultimate Gradle Optimization on M4

How I optimized my Android build time to be faster on M4

Hardware: Apple Silicon architecture (Mac Mini M4 with 16GB RAM):

• ARM64 optimization
• Unified memory architecture
• I/O speeds
• Perfect for parallel workloads

---

Configuration: gradle.properties

org.gradle.jvmargs=-Xmx3072m -XX:MaxMetaspaceSize=1g -XX:+UseG1GC -XX:+AlwaysPreTouch -XX:+UseStringDeduplication
org.gradle.daemon=true
org.gradle.parallel=true
org.gradle.caching=true
org.gradle.workers.max=8
org.gradle.caching.debug=false
org.gradle.configuration-cache=true

What each parameter does:

Parameter	Purpose
-Xmx3072m	3GB RAM for Gradle (sweet spot for M4)
-XX:+UseG1GC	Low-latency garbage collector
-XX:+AlwaysPreTouch	Pre-heats memory for faster access
-XX:+UseStringDeduplication	Eliminates duplicate strings
org.gradle.parallel=true	Uses all M4 cores
org.gradle.configuration-cache=true	caches the configuration phase (important for incremental builds)

---

Results

Configuration	Build Time	Improvement
Default config	~3 minutes	Baseline
Basic memory tweaks	3.045s	98% faster
Adding Parallel GC	5.294s	Wait, worse!
Sweet spot found	1.222s	99.3% faster
With warm caches	0.503s	99.7% faster
Incremental build	0.407s	99.8% FASTER