I was reviewing my posts over the past few days and noticed something that I found quite striking. In this post, I compared the ATMega328 chip, found in the Arduino, against the MSP430 chips found in the TI Launchpad. Note that the MSP430G2211, which is a stripped-down microcontroller with little memory and few peripherals, and is the least expensive of the three chips in the table, is priced the same as the NXP Cortex M0 LPC1111 that I mentioned in this post.
Think about that. For the same price as a 16 bit processor with 2K Flash, 128 bytes of RAM, and a single timer running at 16MHz, you could have a 32 bit processor with 8K flash, 2K RAM, I2C, SPI, a UART, an ADC and four timers running at 50MHz.
Why use such a small 16 bit processor when a much more powerful 32 bit processor can be had for the same price? Honestly, unless you're building a device that needs to run for years on batteries, I don't know why you would. (I haven't reviewed the M0's low power modes to know how they compare against the MSP430.)
Now admittedly, I'm just a hobbyist and don't know much about selection of parts in industry. There may be compelling reasons to choose the MSP430 over the M0. I'm just really struck by the vast difference in two chips that have the same price. (Okay, to be fair, if you get the MSP430G2211 in the QFN package to match the LPC1111, the price does go down to $1.93. But still... that 10 cent difference gets you a huge difference in performance and capabilities.)