Here are my results. I had to trim the hardware and version strings as my boards did not agree with hippy's formatting assumptions. I also added more cpu frequencies and tested from slowest first. This allowed the test to double as a quick check of how much I could overclock the board. Apart from this, I did not modify the code provided by hippy.
I used the official MP versions from the MP website and Pimoroni's "experimental" version from their website.
I could not recreate hippy's Pico results. My Pico seems to run almost twice as fast as his one. Also, my Pico W is generally about 10% slower than my Pico. The MCU's appear to have different steppings, so perhaps their boot ROMs differ.
I have also done some testing with PyStone on both v1.23.0 and v1.24.0. In general v1.24.0 gives a 5% - 10% speedup, except on the Pico W where it gives a 4% - 5% slowdown.
I used the official MP versions from the MP website and Pimoroni's "experimental" version from their website.
I could not recreate hippy's Pico results. My Pico seems to run almost twice as fast as his one. Also, my Pico W is generally about 10% slower than my Pico. The MCU's appear to have different steppings, so perhaps their boot ROMs differ.
I have also done some testing with PyStone on both v1.23.0 and v1.24.0. In general v1.24.0 gives a 5% - 10% speedup, except on the Pico W where it gives a 4% - 5% slowdown.
Code:
PyBoard v1.0:.----------.------------.-------.------------.-------------.| Hardware | Python | MHz | Seconds | Multiplier ||----------|------------|-------|------------|-------------|| STM32F | 1.24.0 on | 125 | 1.945285 | 2 1.9 || | | 133 | 1.945278 | 2 1.8 || | | 150 | 1.620872 | 2 1.9 || | | 168 D| 1.389199 | 3 2.0 || | | 200 | 1.215469 | 3 1.9 || | | 216 | 1.080364 | 3 2.0 || | | 250 | 1.080363 | 3 1.7 || | | 280 | 1.080364 | 3 1.5 || | | 300 | 1.080363 | 3 1.4 || | | 320 | 1.080363 | 3 1.3 || | | 350 | 1.080363 | 3 1.2 || | | 400 | 1.080364 | 3 1.1 || | | 500 | 1.080363 | 3 0.9 || | | 600 | 1.080364 | 3 0.7 |`----------^------------^-------^------------^-------------'MCU is STM32F405RG (Cortex M4 @ 168MHz)D is default frequency before run (set by Thonny?)I think MP is preventing the board running faster than 216MHzRPi Pico:.----------.------------.-------.------------.-------------.| Hardware | Python | MHz | Seconds | Multiplier ||----------|------------|-------|------------|-------------|| RP2040 | 1.24.0 on | 125 D| 1.785654 | 2 2.1 || | | 133 | 1.678249 | 2 2.1 || | | 150 | 1.488043 | 2 2.1 || | | 168 | 1.328608 | 3 2.1 || | | 200 | 1.116059 | 3 2.1 || | | 216 | 1.033391 | 4 2.1 || | | 250 | 0.892853 | 4 2.1 |MCU is RP2040 (RP2-B0 20/24) (Cortex M0+ @ 125MHz)D is default frequency before run (set by Thonny?)This board would not run faster than 250MHzRPi Pico W (WiFi inactive):.----------.------------.-------.------------.-------------.| Hardware | Python | MHz | Seconds | Multiplier ||----------|------------|-------|------------|-------------|| RP2040 | 1.24.0 on | 125 D| 1.918677 | 2 1.9 || | | 133 | 1.803279 | 2 1.9 || | | 150 | 1.598658 | 2 1.9 || | | 168 | 1.427139 | 3 1.9 || | | 200 | 1.198770 | 3 1.9 || | | 216 | 1.110118 | 3 1.9 || | | 250 | 0.959071 | 4 1.9 || | | 280 | 0.856375 | 4 1.9 |MCU is RP2040 (RP2-B2 22/08) (Cortex M0+ @ 125MHz)D is default frequency before run (set by Thonny?)Note: below 216MHz this board is slower than PicoPimoroni Plasma 2350:.----------.------------.-------.------------.-------------.| Hardware | Python | MHz | Seconds | Multiplier ||----------|------------|-------|------------|-------------|| RP2350 | eature/psr | 125 | 1.193519 | 3 3.1 || | | 133 | 1.121759 | 3 3.1 || | | 150 D| 0.994661 | 4 3.1 || | | 168 | 0.888086 | 4 3.1 || | | 200 | 0.745974 | 5 3.1 || | | 216 | 0.690700 | 5 3.1 || | | 250 | 0.596743 | 6 3.1 || | | 280 | 0.532856 | 7 3.1 || | | 300 | 0.497318 | 7 3.1 |MCU is RP2350 (Cortex M33 @ 150MHz)MicroPython is Pimoroni v0.0.9 on 2024-11-06D is default frequency before run (set by Thonny?)Plugging an RM2 module into the SP/CE connector has no effecton the result (WiFi inactive)Statistics: Posted by paulg1 — Mon Nov 11, 2024 3:33 pm