This message will disappear after all relevant tasks have been resolved.

Semantic MediaWiki

There are 1 incomplete or pending task to finish installation of Semantic MediaWiki. An administrator or user with sufficient rights can complete it. This should be done before adding new data to avoid inconsistencies.

This page is a candidate for renaming (move).
The requested new name is: STM32Cube.AI model benchmark .
The supplied reason is: Proposal to avoid capitalizing .
-- Registered User (-) 18:26, 21 September 2021 (CEST).
Wiki maintainers: remember to update the pages that link this page before renaming (moving) it.

This article provides benchmark results for a set of well-known or reference pre-trained Neural Network models. Some STM32 results are candidates for official submission to the MLPerf™ Tiny benchmark from MLCommons™.

Information

STM32Cube.AI (X-CUBE-AI) is a software that generates optimized C code for STM32 microcontrollers and Neural Network inference. It is delivered under the Mix Ultimate Liberty+OSS+3rd-party V1 software license agreement (SLA0048).
The inference time, current and energy measurement process described is not done in a certified laboratory but can be reproduced by any user. The results are average values, which may vary depending on the input data (random data are currently used), the temperature, and the STM32 device itself.
Published data in this article is not contractual.
Copyright STMicroelectronics - All right reserved. Do not publish the following data without written consent of STMicroelectronics

1. Benchmark results

1.1. STM32 High Performance MCUs

STM32 Board	STM32 characteristics	Model Source/Link	Flash (Kbyte)	RAM (Kbyte)	Proc Time (ms)	Cur. (mA)	Energy (mJ) 3.3 V	Version
STM32H723 NUCLEO-H723ZG	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	MobileNet v1 0.25 128 quant tfl source	468 KB	66 KB	49 ms	203 mA	33 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H723 NUCLEO-H723ZG	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	FoodReco quant h5 deriv MobileNet FP-AI-VISION1	132 KB	148 KB	51 ms	197 mA	33 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H723 NUCLEO-H723ZG	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	Person Presence MobileNet v2 128 FP-AI-VISION1	403 KB	197 KB	93 ms	200 mA	62 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H723 NUCLEO-H723ZG	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	Anomaly Detection v0.5 tfl MLPerf™Tiny	265 KB	0.75 KB	1.2 ms	176 mA	0.7 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H723 NUCLEO-H723ZG	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	Key Word Spotting v0.5 tfl MLPerf™Tiny	24 KB	18 KB	11.5 ms	190 mA	7 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H723 NUCLEO-H723ZG	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	Image Classif. v0.5 tfl MLPerf™Tiny	77 KB	49 KB	37 ms	190 mA	23 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H723 NUCLEO-H723ZG	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	Visual Wake Word v0.5 tfl MLPerf™Tiny	214 KB	37 KB	31 ms	198 mA	20 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H735 SMPS STM32H735G-DK	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	MobileNet v1 0.25 128 quant tfl source	468 KB	66 KB	49 ms	97 mA	16 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H735 SMPS STM32H735G-DK	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	FoodReco quant h5 deriv MobileNet FP-AI-VISION1	132 KB	148 KB	51 ms	95 mA	16 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H735 SMPS STM32H735G-DK	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	Person Presence MobileNet v2 128 FP-AI-VISION1	403 KB	197 KB	93 ms	96 mA	30 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H735 SMPS STM32H735G-DK	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	Anomaly Detection v0.5 tfl MLPerf™Tiny	265 KB	0.75 KB	1.2 ms	86 mA	0.34 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H735 SMPS STM32H735G-DK	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	Key Word Spotting v0.5 tfl MLPerf™Tiny	24 KB	18 KB	11.5 ms	97.5 mA	3.74 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H735 SMPS STM32H735G-DK	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	Image Classif. v0.5 tfl MLPerf™Tiny	77 KB	49 KB	37 ms	91 mA	11 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H735 SMPS STM32H735G-DK	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	Visual Wake Word v0.5 tfl MLPerf™Tiny	214 KB	37 KB	31 ms	94 mA	9.6 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H743 NUCLEO-H743ZI	Flash 2 Mbyte RAM 1 Mbyte (512) Freq 480 MHz	MobileNet v1 0.25 128 quant tfl source	468 KB	66 KB	49 ms	203 mA	33 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H743 NUCLEO-H743ZI	Flash 2 Mbyte RAM 1 Mbyte (512) Freq 480 MHz	FoodReco quant h5 deriv MobileNet FP-AI-VISION1	132 KB	148 KB	59 ms	192 mA	37 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H743 NUCLEO-H743ZI	Flash 2 Mbyte RAM 1 Mbyte (512) Freq 480 MHz	Person Presence MobileNet v2 128 FP-AI-VISION1	403 KB	197 KB	108 ms	194 mA	69 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H743 NUCLEO-H743ZI	Flash 2 Mbyte RAM 1 Mbyte (512) Freq 480 MHz	Anomaly Detection v0.5 tfl MLPerf™Tiny	265 KB	0.75 KB	1.4 ms	168 mA	0.78 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H743 NUCLEO-H743ZI	Flash 2 Mbyte RAM 1 Mbyte (512) Freq 480 MHz	Key Word Spotting v0.5 tfl MLPerf™Tiny	24 KB	18 KB	13 ms	196 mA	8.7 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H743 NUCLEO-H743ZI	Flash 2 Mbyte RAM 1 Mbyte (512) Freq 480 MHz	Image Classif. v0.5 tfl MLPerf™Tiny	77 KB	49 KB	42 ms	183 mA	25.6 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H743 NUCLEO-H743ZI	Flash 2 Mbyte RAM 1 Mbyte (512) Freq 480 MHz	Visual Wake Word v0.5 tfl MLPerf™Tiny	214 KB	37 KB	36 ms	189 mA	22 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H747 SMPS STM32H747I-DISCO	Cortex^®-M7 Flash 2 Mbyte RAM 1 Mbyte (0.5) Freq 400 MHz⁽¹⁾	MobileNet v1 0.25 128 quant tfl source	468 KB	66 KB	68 ms	68 mA	15 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H747 SMPS STM32H747I-DISCO	Cortex^®-M7 Flash 2 Mbyte RAM 1 Mbyte (0.5) Freq 400 MHz⁽¹⁾	FoodReco quant h5 deriv MobileNet FP-AI-VISION1	132 KB	148 KB	70.5 ms	69.5 mA	16 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H747 SMPS STM32H747I-DISCO	Cortex^®-M7 Flash 2 Mbyte RAM 1 Mbyte (0.5) Freq 400 MHz⁽¹⁾	Person Presence MobileNet v2 128 FP-AI-VISION1	403 KB	197 KB	130 ms	69.5 mA	30 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H747 SMPS STM32H747I-DISCO	Cortex^®-M7 Flash 2 Mbyte RAM 1 Mbyte (0.5) Freq 400 MHz⁽¹⁾	Anomaly Detection v0.5 tfl MLPerf™Tiny	265 KB	0.75 KB	1.6 ms	64 mA	0.34 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H747 SMPS STM32H747I-DISCO	Cortex^®-M7 Flash 2 Mbyte RAM 1 Mbyte (0.5) Freq 400 MHz⁽¹⁾	Key Word Spotting v0.5 tfl MLPerf™Tiny	24 KB	18 KB	16 ms	70 mA	3.7 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H747 SMPS STM32H747I-DISCO	Cortex^®-M7 Flash 2 Mbyte RAM 1 Mbyte (0.5) Freq 400 MHz⁽¹⁾	Image Classif. v0.5 tfl MLPerf™Tiny	77 KB	49 KB	51 ms	66 mA	11 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H747 SMPS STM32H747I-DISCO	Cortex^®-M7 Flash 2 Mbyte RAM 1 Mbyte (0.5) Freq 400 MHz⁽¹⁾	Visual Wake Word v0.5 tfl MLPerf™Tiny	214 KB	37 KB	43 ms	68.5 mA	9.7 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz	MobileNet v1 0.25 128 quant tfl source	468 KB	66 KB	96 ms	44 mA	14 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz	FoodReco quant h5 deriv MobileNet FP-AI-VISION1	132 KB	148 KB	100 ms	43.5 mA	14 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz	Person Presence MobileNet v2 128 FP-AI-VISION1	403 KB	197 KB	184 ms	44 mA	26 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz	Anomaly Detection v0.5 tfl MLPerf™Tiny	265 KB	0.75 KB	2.3 ms	40 mA	0.3 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz	Key Word Spotting v0.5 tfl MLPerf™Tiny	24 KB	18 KB	23 ms	44.5 mA	3.3 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz	Image Classif. v0.5 tfl MLPerf™Tiny	77 KB	49 KB	72 ms	42 mA	10 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz	Visual Wake Word v0.5 tfl MLPerf™Tiny	214 KB	37 KB	61 ms	43 mA	9 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz AXI 140 MHz⁽²⁾	MobileNet v1 0.25 128 quant tfl source	468 KB	66 KB	97.5 ms	40.6 mA	13 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz AXI 140 MHz⁽²⁾	FoodReco quant h5 deriv MobileNet FP-AI-VISION1	132 KB	148 KB	101 ms	40.2 mA	13 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz AXI 140 MHz⁽²⁾	Person Presence MobileNet v2 128 FP-AI-VISION1	403 KB	197 KB	185 ms	40.4 mA	25 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz AXI 140 MHz⁽²⁾	Anomaly Detection v0.5 tfl MLPerf™Tiny	265 KB	0.75 KB	2.3 ms	36 mA	0.28 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz AXI 140 MHz⁽²⁾	Key Word Spotting v0.5 tfl MLPerf™Tiny	24 KB	18 KB	22.9 ms	40.1 mA	3 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz AXI 140 MHz⁽²⁾	Image Classif. v0.5 tfl MLPerf™Tiny	77 KB	49 KB	72.5 ms	38.4 mA	9.2 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz AXI 140 MHz⁽²⁾	Visual Wake Word v0.5 tfl MLPerf™Tiny	214 KB	37 KB	61.4 ms	39.6 mA	8 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 220 MHz (280) AXI 110 MHz⁽³⁾	MobileNet v1 0.25 128 quant tfl source	468 KB	66 KB	124 ms	28.4 mA	11.6 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 220 MHz (280) AXI 110 MHz⁽³⁾	FoodReco quant h5 deriv MobileNet FP-AI-VISION1	132 KB	148 KB	128 ms	28 mA	11.8 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 220 MHz (280) AXI 110 MHz⁽³⁾	Person Presence MobileNet v2 128 FP-AI-VISION1	403 KB	197 KB	235 ms	28.1 mA	21.8 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 220 MHz (280) AXI 110 MHz⁽³⁾	Anomaly Detection v0.5 tfl MLPerf™Tiny	265 KB	0.75 KB	2.9 ms	25.5 mA	0.24 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 220 MHz (280) AXI 110 MHz⁽³⁾	Key Word Spotting v0.5 tfl MLPerf™Tiny	24 KB	18 KB	29.1 ms	28.7 mA	2.7 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 220 MHz (280) AXI 110 MHz⁽³⁾	Image Classif. v0.5 tfl MLPerf™Tiny	77 KB	49 KB	92.3 ms	26.8 mA	8.2 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 220 MHz (280) AXI 110 MHz⁽³⁾	Visual Wake Word v0.5 tfl MLPerf™Tiny	214 KB	37 KB	78 ms	27.6 mA	7.1 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0

⁽¹⁾ On Cortex^®-M7 core in SMPS mode 400 MHz instead of 480 max in LDO. The Cortex^®-M4 is running on a while(1) infinite loop.

⁽²⁾ The MCU core frequency is set to its maximum 280 MHz, but the HPRE prescaler is set to /2 to provide an HCLK/AXI frequency of 140 MHz instead of the maximum 280 MHz. Note that this setting might impact the rest of the system in a broader application like DMA transfer or ChromeART speed.

⁽³⁾ The MCU core frequency is set to 220 MHz and the HPRE prescaler is set to /2 to provide an HCLK/AXI frequency of 110 MHz to have an optimal power setting with limited impact on the latency.. In particular, the Voltage Scale power regulator can be set to VOS1 instead of VOS0. Note that this setting might impact the rest of the system in a broader application like DMA transfer or ChromeART speed.

For a given STM32 in a fixed configuration, the current consumption is in the same range regardless of the model. it might however vary depending on the complexity and topology of the model. The following table is providing the average current consumption of the model listed in the table above table (excluding the Anomaly Detection model which has a specific topology). These data can be used as a first estimation of the current consumption and the energy consumption of a new model from just the measurement of its inference time. From the average inference time of t second and the average current of i Ampere for a given input voltage of u Volt. The average energy is easily computed as (t x i x u) in Joule.

STM32 Board	STM32H723 550 MHz	STM32H735 SMPS 550 MHz	STM32H743 480 MHz	STM32H747 400 MHz SMPS	STM32H7A3 280 MHz	STM32H7A3 280 MHz AXI 140MHz⁽²⁾	STM32H7A3 220 MHz AXI 110MHz⁽³⁾
Average current (mA)	196	95	191	69	43	40	28

STM32Cube.AI (X-CUBE-AI) can also generate a TensorFlow™ Lite for Microcontroller runtime implementation (version 2.5.0 for STM32Cube.AI v7.0.0). The following table is comparing the TFLm runtime to the X-CUBE-AI runtime.

STM32 Board	STM32 characteristics	Model Source/Link	Runtime	Flash (Kbyte)	RAM (Kbyte)	Proc Time (ms)	Version
STM32H723 NUCLEO-H723ZG	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	Image Classif. v0.5 tfl MLPerf™Tiny	X-CUBE-AI	77 KB	49 KB	37 ms	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H723 NUCLEO-H723ZG	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	Image Classif. v0.5 tfl MLPerf™Tiny	TFL micro	96 KB	53 KB	61 ms	TFLm 2.5.0 Cube IDE 1.7.0
STM32H723 NUCLEO-H723ZG	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	Visual Wake Word v0.5 tfl MLPerf™Tiny	X-CUBE-AI	214 KB	37 KB	31 ms	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H723 NUCLEO-H723ZG	Flash 1 Mbyte RAM 564 Kbyte (432) Freq 550 MHz	Visual Wake Word v0.5 tfl MLPerf™Tiny	TFL micro	325 KB	98 KB	42 ms	TFLm 2.5.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz	Image Classif. v0.5 tfl MLPerf™Tiny	X-CUBE-AI	77 KB	49 KB	72 ms	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz	Image Classif. v0.5 tfl MLPerf™Tiny	TFL micro	96 KB	53 KB	120 ms	TFLm 2.5.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz	Visual Wake Word v0.5 tfl MLPerf™Tiny	X-CUBE-AI	214 KB	37 KB	61 ms	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32H7A3 SMPS NUCLEO-H7A3ZI-Q	Flash 2 Mbyte RAM 1.4 Mbyte (1.18) Freq 280 MHz	Visual Wake Word v0.5 tfl MLPerf™Tiny	TFL micro	325 KB	98 KB	84 ms	TFLm 2.5.0 Cube IDE 1.7.0

1.2. STM32 Ultra Low Power MCUs

STM32 Board	STM32 characteristics	Model Source/Link	Flash Wgt.	RAM Buf.	Proc Time	Cur. (mA)	Energy (mJ) 3.3V	Version
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 160 MHz	MobileNet v1 0.25 128 quant tfl source	468 KB	66 KB	249 ms	10 mA	8.2 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 160 MHz	FoodReco quant h5 deriv MobileNet FP-AI-VISION1	132 KB	148 KB	250 ms	9.1 mA	7.5 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 160 MHz	Person Presence MobileNet v2 128 FP-AI-VISION1	403 KB	197 KB	468 ms	9.4 mA	14.5 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 160 MHz	Anomaly Detection v0.5 tfl MLPerf™Tiny	265 KB	0.75 KB	5.7 ms	9.1 mA	0.17 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 160 MHz	Key Word Spotting v0.5 tfl MLPerf™Tiny	24 KB	18 KB	61 ms	10 mA	2 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 160 MHz	Image Classif. v0.5 tfl MLPerf™Tiny	77 KB	49 KB	164 ms	9 mA	4.9 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 160 MHz	Visual Wake Word v0.5 tfl MLPerf™Tiny	214 KB	37 KB	157 ms	9.1 mA	4.7 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 110 MHz⁽⁴⁾	MobileNet v1 0.25 128 quant tfl source	468 KB	66 KB	361 ms	5.7 mA	6.8 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 110 MHz⁽⁴⁾	FoodReco quant h5 deriv MobileNet FP-AI-VISION1	132 KB	148 KB	363 ms	5.6 mA	6.7 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 110 MHz⁽⁴⁾	Person Presence MobileNet v2 128 FP-AI-VISION1	403 KB	197 KB	681 ms	5.8 mA	13 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 110 MHz⁽⁴⁾	Anomaly Detection v0.5 tfl MLPerf™Tiny	265 KB	0.75 KB	8.3 ms	5.6 mA	0.16 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 110 MHz⁽⁴⁾	Key Word Spotting v0.5 tfl MLPerf™Tiny	24 KB	18 KB	88.2 ms	5.8 mA	1.7 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 110 MHz⁽⁴⁾	Image Classif. v0.5 tfl MLPerf™Tiny	77 KB	49 KB	239 ms	5.6 mA	4.4 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 110 MHz⁽⁴⁾	Visual Wake Word v0.5 tfl MLPerf™Tiny	214 KB	37 KB	229 ms	5.7 mA	4.3 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32L4R5 NUCLEO-L4R5ZI	Flash 2 Mbyte Single Bank RAM 640 Kbyte Freq 120 MHz	MobileNet v1 0.25 128 quant tfl source	468 KB	66 KB	420 ms	23 mA	32 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32L4R5 NUCLEO-L4R5ZI	Flash 2 Mbyte Single Bank RAM 640 Kbyte Freq 120 MHz	FoodReco quant h5 deriv MobileNet FP-AI-VISION1	132 KB	148 KB	433 ms	23 mA	33 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32L4R5 NUCLEO-L4R5ZI	Flash 2 Mbyte Single Bank RAM 640 Kbyte Freq 120 MHz	Person Presence MobileNet v2 128 FP-AI-VISION1	403 KB	197 KB	781 ms	24 mA	63 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32L4R5 NUCLEO-L4R5ZI	Flash 2 Mbyte Single Bank RAM 640 Kbyte Freq 120 MHz	Anomaly Detection v0.5 tfl MLPerf™Tiny	265 KB	0.75 KB	9 ms	22 mA	0.66 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32L4R5 NUCLEO-L4R5ZI	Flash 2 Mbyte Single Bank RAM 640 Kbyte Freq 120 MHz	Key Word Spotting v0.5 tfl MLPerf™Tiny	24 KB	18 KB	104 ms	24 mA	8.3 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32L4R5 NUCLEO-L4R5ZI	Flash 2 Mbyte Single Bank RAM 640 Kbyte Freq 120 MHz	Image Classif. v0.5 tfl MLPerf™Tiny	77 KB	49 KB	255 ms	24 mA	20 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32L4R5 NUCLEO-L4R5ZI	Flash 2 Mbyte Single Bank RAM 640 Kbyte Freq 120 MHz	Visual Wake Word v0.5 tfl MLPerf™Tiny	214 KB	37 KB	260 ms	23 mA	20 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32L4R5 NUCLEO-L4R5ZI	Flash 2 Mbyte Dual Bank RAM 640 Kbyte Freq 120 MHz	MobileNet v1 0.25 128 quant tfl source	468 KB	66 KB	427 ms	24 mA	34 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32L4R5 NUCLEO-L4R5ZI	Flash 2 Mbyte Dual Bank RAM 640 Kbyte Freq 120 MHz	FoodReco quant h5 deriv MobileNet FP-AI-VISION1	132 KB	148 KB	455 ms	24 mA	36 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32L4R5 NUCLEO-L4R5ZI	Flash 2 Mbyte Dual Bank RAM 640 Kbyte Freq 120 MHz	Person Presence MobileNet v2 128 FP-AI-VISION1	403 KB	197 KB	813 ms	25 mA	66 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32L4R5 NUCLEO-L4R5ZI	Flash 2 Mbyte Dual Bank RAM 640 Kbyte Freq 120 MHz	Anomaly Detection v0.5 tfl MLPerf™Tiny	265 KB	0.75 KB	9.7 ms	23 mA	0.73 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32L4R5 NUCLEO-L4R5ZI	Flash 2 Mbyte Dual Bank RAM 640 Kbyte Freq 120 MHz	Key Word Spotting v0.5 tfl MLPerf™Tiny	24 KB	18 KB	111 ms	24 mA	9 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32L4R5 NUCLEO-L4R5ZI	Flash 2 Mbyte Dual Bank RAM 640 Kbyte Freq 120 MHz	Image Classif. v0.5 tfl MLPerf™Tiny	77 KB	49 KB	263 ms	25 mA	22 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32L4R5 NUCLEO-L4R5ZI	Flash 2 Mbyte Dual Bank RAM 640 Kbyte Freq 120 MHz	Visual Wake Word v0.5 tfl MLPerf™Tiny	214 KB	37 KB	270 ms	24 mA	21 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32WB55 SMPS NUCLEO-WB55RG	Flash 1 Mbyte RAM 256 Kbyte Freq 64 MHz	Anomaly Detection v0.5 tfl MLPerf™Tiny	265 KB	0.75 KB	17 ms	9.8 mA	0.55 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32WB55 SMPS NUCLEO-WB55RG	Flash 1 Mbyte RAM 256 Kbyte Freq 64 MHz	Key Word Spotting v0.5 tfl MLPerf™Tiny	24 KB	18 KB	180 ms	11 mA	6.6 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32WB55 SMPS NUCLEO-WB55RG	Flash 1 Mbyte RAM 256 Kbyte Freq 64 MHz	Image Classif. v0.5 tfl MLPerf™Tiny	77 KB	49 KB	457 ms	11 mA	16 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32WB55 SMPS NUCLEO-WB55RG	Flash 1 Mbyte RAM 256 Kbyte Freq 64 MHz	Visual Wake Word v0.5 tfl MLPerf™Tiny	214 KB	37 KB	459 ms	11 mA	16 mJ	STM32Cube.AI 7.0.0 Cube IDE 1.7.0

⁽¹⁾ The MCU core frequency is set to 110 MHz, the Voltage Scale power regulator can then be set to VOS2 instead of VOS1. Note that this setting might impact the rest of the system in a broader application like DMA transfer or ChromeART speed.

The following table is providing the average current consumption of the model listed in the table above table (excluding the Anomaly Detection model which has a specific topology). These data can be used as a first estimation of the current consumption and the energy consumption of a new model from just the measurement of its inference time. From the average inference time of t second and the average current of i Ampere for a given input voltage of u Volt. The average energy is easily computed as (t x i x u) in Joule

STM32 Board	STM32U585 160 MHz	STM32U585 110 MHz⁽⁴⁾	STM32L4R5 120 MHz Single Bank	STM32L4R5 120 MHz Dual Bank	STMWB55 64 MHz
Average current (mA)	9.5	5.7	23.7	24.3	10.8

STM32Cube.AI (X-CUBE-AI) can also generate a TensorFlow™ Lite for Microcontroller runtime implementation (version 2.5.0 for STM32Cube.AI v7.0.0). The following table is comparing the TFLm runtime to the X-CUBE-AI runtime.

STM32 Board	STM32 characteristics	Model Source/Link	Runtime	Flash (Kbyte)	RAM (Kbyte)	Proc Time (ms)	Version
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 160 MHz	Image Classif. v0.5 tfl MLPerf™Tiny	X-CUBE-AI	77 KB	49 KB	164 ms	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 160 MHz	Image Classif. v0.5 tfl MLPerf™Tiny	TFL micro	96 KB	53 KB	315 ms	TFLm 2.5.0 Cube IDE 1.7.0
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 160 MHz	Visual Wake Word v0.5 tfl MLPerf™Tiny	X-CUBE-AI	214 KB	37 KB	157 ms	STM32Cube.AI 7.0.0 Cube IDE 1.7.0
STM32U585 SMPS B-U585I-IOT02A	Flash 2 Mbyte RAM 786 Kbyte Freq 160 MHz	Visual Wake Word v0.5 tfl MLPerf™Tiny	TFL micro	325 KB	98 KB	216 ms	TFLm 2.5.0 Cube IDE 1.7.0

2. Measurement process

On this benchmark only the Machine Learning model inference processing is reported. In a complete application, the sensor acquisition, the data conditioning and pre-processing must also be considered.

The STM32 Board column indicates the STM32 reference and the board used for measurement. By default, the STM32 is configured in maximum performance configuration, so with maximum frequency and especially HCLK / AXI clock at maximal frequency. When a different setting is used it is specified (for instance lower frequency to use a different Voltage Scale or for STM32H7, lower HCLK/AXI frequency). When SMPS is indicated it means that the internal voltage regulator used is the SMPS (Switched-Mode Power Supply) step-down converter instead of the LDO (Linear Voltage Regulator).

The STM32 Characteristics column provides the available internal Flash size, the full internal RAM size and the frequency. The RAM size includes the different kind of memories and banks, TCM, SRAM etc. For the time being, the buffers used by X-CUBE-AI must be placed in a continuous memory area, the maximal RAM size available in continuous area is provided between "()" if not equal to the full size. The frequency indicated is the operating frequency used for the benchmark, so generally the maximal frequency. The only different case is with the STM32H747 Discovery Kit which is operating by default in SMPS power mode and therefore is limited to 400 MHz instead of 480 MHz. Data are rounded to 3 decimals.

The memory footprints are the one reported by X-CUBE-AI using the "Analyze" function (the version of X-CUBE-AI used is mentioned in the table).

The column Model Source/Links indicates the pre-trained ML model and the source, either how it was built / trained or where it can be downloaded. tfl stands for TensorFlow™ Lite .tflite model , h5 stands for Keras .h5 model, quant for quantized models on 8 bits. For FP-AI-VISION1 models, they are located in the package directory: FP-AI-VISION1_V3.0.0\Utilities\AI_resources.

The column Flash reports the Flash occupancy including the model weights, the runtime code generated by X-CUBE-AI to run the neural network and its constants (including the initialized tables).

The column RAM Buf. reports the RAM buffers occupancy, used to store the model activations as well as input and output buffers, and the RAM required by the runtime to inference the model. Note that to gain RAM space the "Use activation buffer for input buffer" and "Use activation buffer for the output buffer" options are selected (through X-CUBE-AI Advanced Settings panel).

The Flash and RAM memory footprints related to the runtime/code execution are computed from the memory map once the X-CUBE-AI validation project for the given model has been built with STM32CubeIDE For X-CUBE-AI runtime, the runtime/code part is computed taking into account NetworkRuntime700_CMx_GCC.a, libc_nano.a and network.o memory footprints. For Tensor Flow Lite for microcontroller runtime, the runtime/code part is computed taking into account all the module used by tflite_micro and libc_nano.a.

The column Proc Time reports the model inference processing time. When the current / energy is indicated, the measure is obtained through the X-CUBE-AI "System Performance" application following the process described on this WiKi article on power measurement. Otherwise the "Validation on target" application is used. In all cases, when generating the application, the selected clock source is always the HSI, X-CUBE-AI is generating first the optimal clock settings and eventually afterwards the clock is set to HSI. STM32CubeMX then autonomously reconfigures the clock settings.

Cur. and Energy is the current and energy computed following the process describe in the WiKi article on power measurement. For STM32 Ultra Low Power microcontrollers, measures are done with the X-NUCLEO-LPM01A power shield as described in the section 4.3.1 "Measure process when current is below 50 mA". For STM32 High Performance microcontrollers measures are done with the Qoitec Otii Arc power analyzer as described in the section 4.3.2 Measure process when current is above 50 mA. In both cases, a 10 s windows is used for averaging) and HSI is selected as clock source.

Accuracy is not reported. X-CUBE-AI is not modifying the DL/ML model topology. The impact on accuracy should be limited. X-CUBE-AI is providing through the "Validation" application a way to measure the accuracy either on x86 or on the target. It can be used to check the eventual impact on accuracy. When running the "Validation on target" application several metrics are computed, one of them is the X-Cross providing error metrics between the original model executed in Python and the C model executed on the target. Random data can be used to compute the RMSE/MAE/L2R errors, however it is recommended to use true data to get the final accuracy. For more details on the metrics, please refer to the X-CUBE-AI Embedded Documentation.

Note that accuracy check is important to compare a float model to a quantize model or when using the Weight compression feature of X-CUBE-AI for float models.

STM32Cube.AI model performances

Contents

1. Benchmark results

1.1. STM32 High Performance MCUs

1.2. STM32 Ultra Low Power MCUs

2. Measurement process