Support for AI acceleration. ESP32-S3 MCU adds vector instructions for accelerating neural network calculations and signal processing. AI developers can use these vector instructions through the ESP- and ESP-NN libraries to achieve high-performance image recognition, voice wake-up and recognition applications. ESP-WHO and ESP-Skainet also support this feature. ESP32-S3 provides a complete security mechanism and protection measures for IoT devices to prevent various malicious attacks and threats. It supports flash encryption based on -XTS algorithm, secure boot based on algorithm, digital signature and HMAC. Wi-Fi + Bluetooth 5 (LE): ESP32-S3 integrates 2.4 GHz Wi-Fi (802.11 b/g) and supports 40 MHz bandwidth; its low-power Bluetooth subsystem supports Bluetooth 5 (LE) and Bluetooth Mesh, which can achieve long-distance communication through Coded PHY and broadcast extension. It also supports 2 Mbps PHY to increase transmission speed and data throughput. ESP32-S3's Wi-Fi and Bluetooth LE radio performance are excellent and can work stably even at high temperatures. Rich IO interfaces: ESP32-S3 has 45 programmable GPIOs and common peripheral interfaces such as SPI, I2S, I2C, PWM, , ADC, UART, SD/MMC host controller and TwAITM controller. 14 of the GPIOs can be configured as capacitive touch inputs for HMI interaction. In addition, ESP32-S3 is equipped with an ultra-low power coprocessor (ULP), which supports multiple low-power modes and is widely used in various low-power application scenarios. Note: Due to different production batches, the printing screen printing/appearance is different, which does not affect the use. Support AI: Qwen2.5, DeepSeekV3, Doubao (1.5 Professional Edition), QwenMax2.5 Product design updates May result in slight differences ,If you mind, Please do not order. It's DIY kits, nee to by yourself. Specifications:DIY Kit Colour:White & Black Material:PCB+ components+plastic Package Contents: 1 * ESP32-S3-AI Kit Only the above package content, other products are not included. Note: Light reflection and different displays may cause the color of the item in the picture a little different from the real thing. The measurement allowed error is +/- 1-3cm.