MS All Models Studio M1 Max vs. M4 over bc

[apostolosdt]

For what it's worth.

I ran a simple practical comparison of the new M4 over a Studio Max M1 thru Terminal bc runs. I compared the run times of their CPUs with the bc calculation of 123456789^1234567 and I got these results:

MBP M4 cpu 26.543 sec, vs. M4 iMac cpu 27.668 sec, vs. M1 Max cpu 37.072 sec, all total.

All three machines were of the base type. By the way, the Studio M1 Max needed a total of 24 min 3.71 sec to evaluate the 123456789^12345678. Well, yeah.

 

[mattmower]

On my M2 Max studio this runs in 63-65s.

I'm a little puzzled at your result of "37.072s" for an M1 Max cpu.

 

[apostolosdt]

I assume you run the command (or analogous) "time echo 123456789^1234567 | bc".

I don't have access to my M1 Max right now, but the same run on a Mac mini 2018 (3 GHz 6-core i5) has just given this:

which is reasonable for this kind of Mac.

 

[chars1ub0w]

On my M2 Max studio this runs in 63-65s.

I'm a little puzzled at your result of "37.072s" for an M1 Max cpu.

OP is right.

~$ sysctl -n machdep.cpu.brand_string
Apple M1 Max
~$ time echo 123456789^1234567 | bc
...
real 0m37.261s
user 0m36.753s
sys 0m0.373s

~$ sysctl -n machdep.cpu.brand_string
Apple M3
~$ time echo 123456789^1234567 | bc
...
real 0m32.355s
user 0m32.128s
sys 0m0.143s

 

[apostolosdt]

On my M2 Max studio this runs in 63-65s.

I'm a little puzzled at your result of "37.072s" for an M1 Max cpu.

I wouldn’t put much weight on that kind of “test”. The whole thing started as sort of fun when I visited a newly open Apple Store nearby. Since then, it has become a habit of mine with every new Mac.

But this thread made me want to dig a bit deeper into what that Terminal run measures. Ok, looks like it tests CPU’s speed, still—does anyone know better?

Interested to hear.