Eloquent Arduino Pits Three Popular Dev Boards Head-to-Head to See Which Comes Out on Top for TinyML
Simone Salerno puts the Arduino Portenta H7, Teensy 4.0, and STM32 Nucleo H743ZI2 through their paces for on-device machine learning.
Eloquent Arduino's Simone Salerno has published benchmarks pitting three popular development boards against each other for their performance at running on-device machine learning — TinyML.
"A few days ago," Salerno writes by way of introduction, "I asked a poll on my Twitter for who do you think would be the fastest board for TinyML among Arduino Portenta H7, Teensy 4.0, and STM32 Nucleo H743ZI2. Both Portenta and Nucleo ranked on par at first position, leaving Teensy behind. All of them share an ARM Cortex M7 CPU, but which one is the winner?"
To find out, Salerno came up with a series of tests part of a head-to-head benchmark. The benchmark itself measures inference time for everything from digits and gesture phase segmentation to gas sensor array drifting and sport activity, across a range of approaches. The lower the time, the faster the microcontroller at that task.
With these measurements in hand, Salerno takes a look at how often each of the three microcontrollers ranked in various positions, then produced a "global ranking" to see which of the three came out on top. The result: the Teensy 4.0 topped the league.
"Teensy 4.0 is a clear winner, since it's the fastest on most of the cases. Is it surprising," Salerno asks. "Not at all, if you consider it is clocked at 600 MHz vs 480 MHz of Arduino Portenta H7 and STM32 Nucleo H743ZI2. STM32 Nucleo H743ZI2 is fast on Decision Tree, Random Forest and Gaussian NB, but not so fast on the other classifiers. Arduino Portenta H7 is slower than Teensy 4.0, but generally faster than STM32 Nucleo H743ZI2, ranking most of the times at position 1st and 2nd."
Salerno's full benchmark write-up is available on the Eloquent Arduino website.