You are right about the ESP32 having 448KB of ROM.
The ESP32 i a very interesting product, but as told it has some development issuesā¦
Could it happen to support Particleās tool setā¦? You do need to make some new boards from time to time, right? 
It would be really useful to have the cloud functions for it, if not the whole Particle tool set!
I do not know the Tensilica LX6 (processor) but, well, its probably very different from the Cortex, meaning much work?
Weāre definitely keeping an eye on the ESP32.
Generally speaking we prefer to choose chips and modules from vendors with a strong history, a great reputation, and good support. The BCM43362 in the Photon is from Broadcom/Cypress, and the module is manufactured by one of the largest Wi-Fi module manufacturers in the world (they supply Wi-Fi modules for the iPhone).
All that said, weāre keeping our eye on the ESP32, in case it becomes stable and well-supported enough for us to recommend it.
In the meantime, if anyone has any interest in creating a Compound for the ESP32 (https://www.particle.io/products/hardware/compounds), weād be happy to support that effort!
The ESP32 (the chip inside the ESPWROOM32) contains two processors, one of them has two cores, for a total of 3 cores in this bitch. Inside the ESP32 is 520kB SRAM and 448kB of ROM.
Externally supported is ā4 x 16MBā of QSPI flash and SRAM, up to 16MB of flash are mapped to the CPU code space, up to 8MB of SRAM are memory mapped to the CPU data space. The thing is under $4, adding flash would increase the price a ton considering how cheap the whole module is, and not everyone needs the flash so itās supported rather than included.
Other features also include:
802.11 b/g/n/d/e/i/k/r (802.11n up to 150Mbps) WIFI, Bluetooth 4.2 BR/EDR & BLE
- *Class 1, 2, and 3 transmitter. In english that means it has 100, 10, and 1 meter wireless range.
- *Station/softAP/SoftAP+Station/P2P, WPA/WPA2/WPA2-Enterprise/WPS, IPv4, IPv6, SSL, TCP/UDP/HTTP/FTP/MQTT, SDIO/SPI/UART/I2S/I2C, Ethernet, and SD-Card stuffs.
Encryption: AES, RSA, ECC, SHA
CVSD (voice modulation) & SBC (Bluetooth audio codec)
PWM, IR Remote stuffs, and Low noise analog Pre-amp?
Hall effect, capacitive touch, and temperature sensor
Yo, Particle ā Pay particular attention to that part where the ESP32 has IPv6 support.
I asked you guys for that in 2014 and you said it was a high priority in 2015.
Hello⦠Is this thing on? (taps microphone)
As to strong history and great reputation, well, the ESP8266 did pretty well and everything Iāve read about the ESP32 promises great things. Iāve been holding off hoping to stick with Particle waiting for IPv6 support, but looks like Iām going to have to move forward and try the ESP32 out myself.
Acknowledging your multiple posts on IPv6 support, I just wanted to mention that the new Argon uses a ESP32 for its networking. Might be of interest to you.
Might I also add that youāre the only one (in recent months/years?) that has asked for IPv6 on these forums. As such, it doesnāt feel like a high priority issue, or more people wouldāve asked for it since. Thatās not to say it isnāt or shouldnāt be, thatās just what Iāve seen.
To add to that, IPv6 will be explicitly discussed with Particle soon and an open GitHub issue about the pending support is open for contributing comments.
āSpammingā the forum isnāt really the most effective way to increase the topicās priority - there are better ways and GitHub issues are one of them.
Folloging the comments in there may give you a feeling about the reasons behind the delay
@owendelong, the next generation of Particle devices will be using Thread meshing, a protocol built on 6Lowpan, based on IPv6.
There is no question that IPv6 is a must for new devices. However, IPv4 is still quite viable for a large percentage of Photon applications. Iām not sure at this time if the investment of work to implement IPv6 on the Photon (or Electron) is worthwhile. This would be a question for the folks at Particle - @rickkas7?
Iāve just reviewed the primary forum post about IPv6. Iām not the only one who asked, just the most persistent. The fact that others have given up and moved on to other platforms (some even stated as much) should, IMHO, provide more incentive to be competitive, I would think.