Thank you so much, mbotmiller, for your reporting on the M1 Mac Mini! I just got one and will be trying it out.

You're very welcome, nickoli - it was fun to run those tests and monitor electricity usage while doing so and share that information with the community! Hope it goes well with your M1 Mac mini.


I'm with you jmenard - I'll be very curious when that happens to take another look at the performance and efficiency numbers. I'm guessing it may be a while though (months to years?) before the underlying science applications that a number of the projects use are rewritten to run natively on Apple Silicon processors running macOS. I'll also be curious whether the GPU in the M1 chip is supported at all if/when the OpenPandemics project goes production with GPU work units - that may happen sooner than the other underlying science applications being rewritten for Apple Silicon.

I wanted to share a quick update on an experiment I’ve been trying with my M1 Mac mini with 8GB of memory. Given that the M1 chip architecture has four high‑efficiency cores and four high-performance cores, I wondered what would happen if I limited WCG to only run on the four high-performance cores (via limiting the BOINC agent to only use 50% of the cores in the system - macOS should automatically assign the four WCG threads to the four high-performance cores).

The results I have observed when doing that are quite promising, achieving comparable or slightly elevated output compared to when using all eight cores. For the past seven days, my base M1 Mac mini has averaged both 44,800 points and 152 results per day while exclusively running the Mapping Cancer Markers project on only the four high-performance cores (see https://imgur.com/a/4vHaOAL for a screenshot of that table). In addition, in a sample size of 75 valid results that I analyzed, the average completion time was 0.62 hours per result and ranged from 0.2 hours on the shortest end to 1.03 hours on the longest end. Moreover, these results were achieved while I was also using this same Mac mini for ~7-9 hours of work on 4 of the 7 days (consisting mostly of video meetings using the M1-optimized version of Zoom, email, spreadsheets, etc.) For owners with a M1 model that has 8GB of memory, another benefit of only using four cores is reducing WCG memory usage by half.

I also wanted to share a nifty command I came across to see more detailed information about CPU utilization and power usage on macOS: sudo powermetrics -s cpu_power

Here is a screenshot of what the output of that terminal command looks like on a M1 Mac mini: https://imgur.com/a/qOuHYDv

I’ve continued to be pleased with the performance of the M1 chip. The efficiency is still outstanding with the total average system power draw of ~25 watts resulting in 0.6 kWh used per day. Based on this past week's worth of MCM data, roughly 74,500 points are generated per kWh for that project.

Found Yaantra repair services through word of mouth and till date its the best.
They offer amazing services at affordable prices.
If you do find any better service do let me know.

IPhone 11 Pro Max MOTHERBOARD REPAIR
http://phonecash.yaantra.com/apple/iphone-11/

I'm also about to get an M1 Mini, but as a "daily driver" replacing an old MacBook Pro. While I hope to do some development on it as well as my normal day to day stuff, it will definitely be running BOINC the rest of the time. Whether WCG's app requires more RAM or not, because it is not upgradable later, I intend to spring for the extra 8GB - "you can never have enough RAM." Storage, eh - can always add external drives. But 8GB seems REAL tight to me, even for a machine that won't be heavily loaded.