TI MSP430 vs. Microchip PIC: low-power showdown
![]() |
0.0 (0) |
The TI MSP430 and Microchip PIC target many of the same apps, and the competition between these parts is intense. Both TI and Microchip claim to have the world’s lowest-power parts, and both companies have produces app notes, videos, and web pages to prove their point. So who really has the lowest-power parts?
On the surface, Microchip seems to have the upper hand. Its parts boast standy power as low as 0.03 μA @ 3 V, compared to 0.1 μA @ 3 V for TI. Microchip also has impressively low active currents, with some parts running below 50 µA/MHz. In contrast, the MSP430 family only gets down to 160 μA/MHz.
Dig deeper, though, and it’s not so obvious that Microchip is winning the power war. On the question of standby power, the MSP430 has an always-on brown-out reset (BOR) that bumps up its standby power numbers. In contrast, the PIC devices achieve their ultra-lower power numbers by disabling the BOR. TI thinks that this is an unrealistic scenario, and that the MSP430 comes out ahead when you factor in the PIC BOR power. Not to be outdone, Microchip responds that they win even if you factor in BOR.
As if this wasn’t confusing enough, you also have to factor in wake-up time to get a complete picture of sleep power. Once again the numbers are all over the place, with both companies claiming the advantage. (See for example the Microchip page that shows a 3 µs MSP430 wake-up time instead of the 1 µs TI claims.) Good luck sorting it all out. (And I haven’t even attempted to factor in the power savings TI gets from its smart peripherals!)
Things are just as confusing on active power. Direct power comparisons are difficult because the PIC is an 8-bitter, while the MSP430 is a 16-bit machine. That mean the MSP430 can complete a given task in fewer cycles (or a lower clock speed). On the other hand, the architectural differences between the two machines is not as big as it appears on the surface. The latest PICs use an enhanced core with 14-bit instructions. This core has 49 instructions, compared to 27 on the MSP430. So it’s not obvious that the MSP430 has a huge architectural advantage over the latest PIC parts. (I realize that counting instructions is a pretty lousy way to gauge performance, but I can’t find any worthwhile benchmarks for the two families. Anybody know where I can find some solid data?)
It’s worth noting that this same concept extends to code density. The MSP430 has a huge advantage in code density over traditional 8-bit architectures, but it’s not clear that it has much advantage over the enhanced PIC core.
The take-away here is that both TI and Microchip have parts with impressively low power, but picking a winner is surprisingly difficult. Much depends on the specific of your application. Do you need the MSP430’s always-on BOR? Are you doing 16-bit math? Etc. I wish I had a more definitive conclusion, but I’m afraid you’ll have to do your own homework!
User reviews
To write a review please register or login.
|
Category: TB-Blog
|
|
Category: TB-Reviews
|
|
Category: TB-Articles
|
|
Category: TB-Blog
|
|
Category: TB-Blog
|
|
Category: TB-Blog
|
|
Category: TB-Blog
|
|
Category: TB-Blog
|
|
Category: TB-Reviews
|
|
Category: TB-Reviews
|
|
Category: TB-Reviews
|
|
Category: TB-Reviews
|
|
Category: TB-Reviews
|
|
Category: TB-Reviews
|
|
Category: TB-Reviews
|
|
Category: TB-Reviews
|
|
Category: TB-Reviews
|
|
Category: TB-Blog
|
|
Category: TB-Blog
|





