Comparing the speed of a bunch of Arduino boards

Go To Last Post
3 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I thought the AVRFREAKS crowd would like a couple of recent blogs we posted. Bonus points if you spot AVRMan!

The first one was a match-up between that AVR Mega and the PIC16 here https://www.microforum.cc/blogs/entry/47-comparing-raw-pin-toggling-speed-avr-atmega4808-vs-pic-16f15376/

 

The second one was a comparison between various Arduino boards here https://www.microforum.cc/blogs/entry/48-comparing-raw-pin-toggling-speed-between-arduino-platforms/

 

We are planning on doing a comparison of interrupt latency next, and who knows what after that :)

Take a look around the site as well, may just be something else that interests you.

An old man that likes to play with Microcontrollers. I also blog at https://www.microforum.cc/blogs/

Last Edited: Thu. Dec 19, 2019 - 06:01 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

It might make more sense to "normalize" to max chip frequency, rather than to clock rate.  Micros with more cycles per instruction also tend to have higher clock rates overall.

 

Otherwise, there's not really anything new there.

 

Arduino's "digitalWrite()" has so much crap in it, and is so inconsistently optimized between different platforms, that it's almost worthless to measure. :-(

(Did you notice that it has different speed on different pins, on the standard AVR boards?)

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What do you mean by max chip frequency? I ask because that was what we tried to do. We did not want to use the instruction clock as the PIC using 4 clocks per instruction would look faster even if it was running slower in reality.

On the other side we also wanted to know if the cpu’s were running at the same clock (oscillator) frequency how do they compare?

Datasheets always list max oscillator speed so if you have a normalized value you can scale the performance to the actual device you are using. Some PIC’s can only run to 20MHz and others can run up to 64MHz so it was important for us to be able to compare devices taking clock speed into account.

It was interesting on the Teensy that the digitalwrite only added 8 cycles to the loop though! This was better than we expected.

An old man that likes to play with Microcontrollers. I also blog at https://www.microforum.cc/blogs/