M1 vs i9: Jeff Geerling’s Kernel Compile Test Shows 30% Faster Build Times

Background

Jeff Geerling needed to rebuild the Raspberry Pi OS Linux kernel for the aarch64 (ARM64) architecture. To compare compilation performance across different Mac hardware, he compiled the same kernel source on three machines using identical build settings.

Test Machines

Intel i9 16‑inch MacBook Pro – 3,000 USD, fan‑cooled, high noise and heat.

Apple M1 Mac mini – comparable price, fan‑cooled.

Apple M1 MacBook Air – fan‑less, relies on thermal throttling.

Compilation Procedure

The kernel source was fetched from the official Raspberry Pi OS repository and built with the standard make workflow for aarch64. All three systems used the same make flags, toolchain version, and number of parallel jobs (matching the CPU core count). No additional optimisations or caching were applied.

Results

M1 Mac mini : completed the build in under 9 minutes . Fan noise was negligible and the chassis temperature remained low.

M1 MacBook Air : took about 1 minute longer (~10 minutes) because the lack of a fan caused the M1 to throttle slightly.

Intel i9 MacBook Pro : required approximately 12.5 minutes , making it roughly 30 % slower than the M1 devices.

Geerling noted that cross‑compiling x86 code on ARM incurs inherent disadvantages, yet the native ARM performance of the M1 still outperformed the Intel CPU.

Additional Observations

When an external 4K monitor was connected via HDMI, the M1 Mac mini experienced intermittent black‑outs lasting 2–4 seconds every five minutes. Using a CalDigit TS3 Plus dock with a DisplayPort connection eliminated the issue; the monitor automatically powered on after the Mac woke from sleep.

Conclusion

The test demonstrates that Apple‑silicon Macs provide faster compilation times, lower power consumption, and quieter operation for native ARM workloads compared to high‑end Intel Macs. While the M1 is not without quirks (e.g., HDMI display instability), its overall efficiency makes it a superior choice for development tasks such as rebuilding the Raspberry Pi OS kernel.