I had to pause some of my work getting a current-gen AMD graphics card running on the Pi 5 and testing a 192-core AmpereOne server to quickly post on the M4's efficiency.
I expected M4 to be better than M1/M2 (I haven't personally tested M3), and I hoped it would at least match the previous total-system-power efficiency king, a tiny arm SBC with an RK3588 SoC... but I didn't expect it to jump forward 32%. Efficiency gains on the Arm systems I test typically look like 2-5% year over year.
The M4 mini I just bought reaches 6.74 Gflops/W on the HPL benchmark.
UPDATE 2024-11-20: I re-ran all my tests without my 10 Gbps Ethernet plugged in, and the efficiency is even better. I also re-tested LLM performance, and made this quick video:
For the latest benchmark results, check my sbc-reviews M4 Mac mini issue.
I can get 283 Gflops at 42W, versus 264 at 66W on my M1 Max Mac Studio (for a round 4.00 Gflops/W).
The chip isn't the fastest at everything, but it's certainly the most efficient CPU I've ever tested. And that scales down to idle power, too—it hovers between 3-4W at idle—which is about the same as a Raspberry Pi.
This is total system power draw, too—not just the CPU.
And the system I bought includes 10 Gigabit Ethernet and 32 GB of RAM; most systems I've used consume 4-6W just running the 10 GbE controller!
In 1.25U of rack space, you could run three Mac minis, idling around 10W, giving almost a teraflop of CPU performance. (Not to mention there's a fast GPU/NPU, 10 GbE, and tons of high speed Thunderbolt IO in the back.)
If only they didn't put the power button on the bottom.
You can check out all my top500-benchmark results on GitHub, and follow my progress testing out the M4 mini here.
I haven't tested an M4 Pro Mac mini yet, so I'm not sure if the efficiency is any better or worse.