What are these options for? They don't seem to be documented yet.

@stanely You are my hero - thanks for figuring this out!

@rashedtalukder Super! An officially supported MicroPython package for the Core2 for AWS would be amazing.

Has anyone been able to port this to the Core 2? I tried but the SD Card library seemed to be incompatible.

What service powers the voice recognition for the Atom Echo in the example:

https://docs.m5stack.com/#/en/quick_start/atom/atom_echo_quick_start

Is it Baidu?

If anyone else is trying to do this you'll need to fix your partition table to take advantage of the larger memory. The steps how to do this are:

1. Go to PIO Home in Visual Studio Code
2. Click on +New Project
3. Name it m5core2 or similar. Choose M5Stack Core ESP32 for the board.
4. Leave Framework set to Arduino.
5. Open your platformio.ini file and replace the text with this:

[env:m5stack-core-esp32]
platform = espressif32
board = m5stack-core-esp32
framework = arduino
; [[[for macos]]]
;upload_port = /dev/cu.SLAB_USBtoUART
; [[[fix PSRAM size and you won't need have this file]]]
board_build.partitions = default_16MB.csv
build_flags =
    -DBOARD_HAS_PSRAM
    -mfix-esp32-psram-cache-issue

6. Copy the 16MB partition table file into your src directory. You can download this file here:

https://github.com/espressif/arduino-esp32/blob/master/tools/partitions/default_16MB.csv

7. In the 'lib' directory you will need three libraries.
   FastLED - https://github.com/FastLED/FastLED
   M5Core2 - https://github.com/m5stack/M5Core2
   ArduinoECCX08 - This one is in a zip file in the M5Core2 library repo:
   https://github.com/m5stack/M5Core2/blob/master/examples/core2_for_aws/ArduinoECCX08.zip

Your lib directory should look like this when you're done:

8. Now copy everything from the example FactoryTest directory in the M5Core2 library into your src folder. The directory files can be viewed here:

https://github.com/m5stack/M5Core2/tree/master/examples/core2_for_aws/FactoryTest

9. Rename FactoryTest.ino to main.cpp

10. Edit main.cpp and comment out these two lines with //. These tests will fail if you don't have an SDCard or something plugged into the IO port.

11. Connect your M5Core2 and build/upload. You can do CTRL-Shift-P to find the Platformio:Upload option.

12. Play with the cool factory demo!

Thanks to jokercatz for the tip on fixing the platform:
http://jokercatz.blogspot.com/2020/11/m5stack-core2-build-from-platformio.html

It looks like these LEDs are not supported in UIFlow, and the NeoPixel micropython library didn't work.

I was able to build new firmware using PlatformIO and flash the FactoryTest application here:

https://github.com/m5stack/M5Core2/tree/master/examples/core2_for_aws/FactoryTest

This has an example of using the FastLED library to control these RGB LEDs.

I flashed my M5Stack Core2 AWS edukit version with UiFlow.

The docs indicate that the SK6812 LEDs are controlled with GPIO25:

SK6812-LED
ESP32 Chip GPIO25
SK6812-LED DATA

I tried this code to test them out, and while the code runs without error, the lights never come on:

from machine import Pin
from neopixel import NeoPixel

pin = Pin(25, Pin.OUT)  # Pin 25 to drive NeoPixels
print("Creating NeoPixel object")
np = NeoPixel(pin, 10)  # create NeoPixel driver on GPIO25 for 10 pixels
np[0] = (255, 0, 0)     # set first pixel red, full brightness
np[1] = (0, 128, 0)     # second pixel green, half brightness
np[2] = (0, 0, 64)      # third pixel blue, quarter brightness
np[3] = (255, 255, 255)
np[4] = (255, 255, 255)
np[5] = (255, 255, 255)
np[6] = (255, 255, 255)
np[7] = (255, 255, 255)
np[8] = (255, 255, 255)
np[9] = (255, 255, 255)
np.write()              # write data to all pixels
print("done")

Anyone have ideas how to access these LEDs with micropython?

In the meantime if you want to play around with AWS and MQTT I wrote up this how to guide for setting it up in UIFlow:

https://community.m5stack.com/topic/2688/mqtt-config-for-aws-iot-core-solved

You can both publish and subscribe to topics if you follow those steps. You do need to flash your device with the UIFlow image though.

I am also stuck trying to get this to work. Some research reveals that there may be a ca_certs parameter required to include Amazon's root CA. I've tried it like this but no luck:

while True:
  output.setText('Setting up mqtt')
  m5mqtt = M5mqtt('m5stick', 'a2zey9c7ts6fdf-ats.iot.us-west-2.amazonaws.com', 1883, '', '', 300, ssl = True, ssl_params = {'key': "/flash/res/priv.key", 'cert': "/flash/res/cert.pem", 'ca_certs': "/flash/res/ca.pem"})
  m5mqtt.subscribe(str(''), fun__)
  output.setText('starting mqtt')
  m5mqtt.start()
  output.setText('publishing hello world')
  m5mqtt.publish(str('test'),str('hello world'))
  output.setText('sent hello world')
  wait(1)
  wait_ms(2)

Can anyone from m5stack confirm whether ca_certs is a supported ssl parameter?