Having used the Arduino prototyping platform (a loose combination of specific pieces of somewhat open/free hardware and a more open/free software stack) for some time for educational and tinkering purposes in my local hackerspace (geeklabs.dk) I have seen and studied the Arduino UNO hardware and lots of its “clones/compatibles/knockoffs” and their common MCU (MicroController Unit);
Atmel ATmega328P
I had begun wondering what the P in the microcontroller model name actually meant. So here is an attempt to decode the Atmel megaAVR chip numbering system. The other existing AVR based series UC3, tinyAVR, XMEGA, Battery & Automotive, will probably employ similar naming schemes.
The remainder of the product name following “ATmega” expresses the available flash memory and the approximate pin count of the package in an integer and optionally other features as either integer or letter (like the initial wondering of P in 328P above).
Starting with the integer, it is a concatenation of two separate integers encoding the flash size and pin count as defined below. The division of the two is non-ambiguous leaving some interpretation to be done.
1st integer: onboard flash size
8 = 4 KiB
8 = 8 KiB
16 = 16 KiB
32 = 32 KiB
64 = 64 KiB
128 = 128 KiB
256 = 256 KiB
2nd integer: total pin number
(none) = standard pin count (differs)
8=28/32-pin
4= 40/44/49-pin
5= 64-pin
0= 100-pin
Suffix (char or integer), multiple possible
P = picoPower (max. consumption 9mA@8MHz,5v vs. 12mA@8Mhz,5v for non-P)
9=LCD controller
U2 = USB controller
U4 = USB controller
A = ?
Exceptions
Note that some of the product names are completely void of these rules. Others employ different numbering but still with a familiarity to the above.
An example:
ATmega6490A: 64KB flash, 100-pin, LCD Controller
Sources
