We might need new types of chip designs for the internet of things


If we’re going to embed chips into roads, bury them in fields and even slap them on produce at the grocery store, we’re going to have to dump the battery. They are expensive, bulky, require changing and can also leak toxins into the environment. And while many people are focused on how to best to power these future sensors, the chip architecture firm ARM(s armh) is thinking about how to build chips for sensors that harvest their own energy.

Maybe these sensors are gathering power from RF signals in the air, kinetically, or even through a chemical reaction that they are built to recognize and report. No matter how they gather the energy, if they can’t store it then the way the chips use that power may have to change. At least that’s what Mike Muller, the CTO of ARM, told Peter Clarke of EETimes last week. In this story he explains how ARM is rethinking chip design for specific internet of things applications:

“Normally, the best strategy is to do processing as fast as possible and then go to sleep for as long as possible — get in and get out,” Muller said. “But for energy scavenging, it can be different.” In these cases, it may be best to power up and get the packet out at minimum energy as soon as enough energy has been harvested. “And it turns out the design tradeoffs are different.”

The story explains how the design considerations mean that an energy-harvesting chip might operate at near the power threshold required to flip a transistor on or off as opposed to the norm of powering something all the way up and down. ARM is working on this technology with its foundry partners to ensure that such a different design could be manufactured.

Should this design win converts, it would take ARM’s reputation as a power-sipping core to entirely new levels. Of course, these chips would be slower than page loads on a 2G network, but perfectly tailored for delivering little bits of environment information.


Andreas The MCU guy

The compute fast, sleep long strategy has a serious limitation in energy harvesting applications, namely peak current consumed during the computation period. It is limited what you solid state battery or large cap can deliver.

your on time and computation needs will be very determined by what protocol we (as a community) select as the IoT protocol, WiFi( requiring a TCP/IP stack) that talks directly to a WiFi access point, or smaller RF protocols that have to go through a hub of some kind before they reach the internet.

The ideal solution for energy harvesting would be a fully asynchronous MCU, but that does not exist yet. In the mean time the race is on for both low power micros and low power trancievers


Clayton Plymill, LEED AP

It would be slower true, but you really don’t need most environmental data every second. I like the way they are thinking!


Isnt this similar to technology from Passif (acquired this month by Apple)

Comments are closed.