MicroPython • Re: Benchmark Pico 2

Thanks, Hippy, for explaining why there is a difference between time and utime.

After some messing, I did test on both architectures with no OC (I don't know how) on a stock pico2. Just wanted to get an idea which Micropython build offers what performance. Not sure what viper is but that's the area of difference.

MicroPython v1.24.0 preview 321

./run-perfbench.py -s ~/pico2-m33.txt ~/pico2-RISC-V.txt
diff of scores (higher is better)
N=150 M=96 /pico2-m33.txt -> /pico2-RISC-V.txt diff diff% (error%)
bm_chaos.py 436.39 -> 273.43 : -162.96 = -37.343% (+/-0.05%)
bm_fannkuch.py 97.22 -> 85.25 : -11.97 = -12.312% (+/-0.04%)
bm_fft.py 3835.41 -> 2850.05 : -985.36 = -25.691% (+/-0.02%)
bm_float.py 6516.82 -> 3699.18 : -2817.64 = -43.236% (+/-0.11%)
bm_nqueens.py 4733.36 -> 4205.95 : -527.41 = -11.142% (+/-0.04%)
bm_pidigits.py 1228.99 -> 953.95 : -275.04 = -22.379% (+/-0.08%)
bm_wordcount.py 72.67 -> 49.88 : -22.79 = -31.361% (+/-0.03%)
core_import_mpy_multi.py 576.54 -> 399.31 : -177.23 = -30.740% (+/-0.06%)
core_import_mpy_single.py 107.55 -> 76.02 : -31.53 = -29.317% (+/-0.20%)
core_locals.py 58.62 -> 45.50 : -13.12 = -22.381% (+/-0.01%)
core_qstr.py 215.84 -> 210.15 : -5.69 = -2.636% (+/-0.03%)
core_str.py 31.49 -> 30.03 : -1.46 = -4.636% (+/-0.04%)
core_yield_from.py 382.42 -> 378.25 : -4.17 = -1.090% (+/-0.01%)
misc_aes.py 668.20 -> 568.13 : -100.07 = -14.976% (+/-0.10%)
misc_pystone.py 2564.50 -> 2059.42 : -505.08 = -19.695% (+/-0.07%)
viper_call0.py 575.30 -> 606.04 : +30.74 = +5.343% (+/-0.01%)
viper_call1a.py 561.09 -> 589.37 : +28.28 = +5.040% (+/-0.00%)
viper_call1b.py 455.42 -> 487.74 : +32.32 = +7.097% (+/-0.00%)
viper_call1c.py 462.44 -> 489.33 : +26.89 = +5.815% (+/-0.00%)
viper_call2a.py 550.79 -> 578.02 : +27.23 = +4.944% (+/-0.01%)
viper_call2b.py 403.92 -> 432.83 : +28.91 = +7.157% (+/-0.00%)

Statistics: Posted by hartsr — Sun Oct 13, 2024 8:20 am

---