ARM made today’s announcement of the new “PlasticARM” prototype along with semiconductor firm PragmatIC. ARM says this new 32-bit chip is not based on existing semiconductor designs, instead relying on metal-oxide thin-film transistor technology on a flexible substrate. Some of the components on this new plastic chip are familiar — there’s a block of RAM (just 128B of it) and an equally anemic allotment of ROM storage. There’s also a Cortex-M CPU core, which supports the older but still serviceable ARMv6-M architecture.
The M0 CPU core is a low-power part, but it’s not new to the plastic chip. The innovation is the flexible substrate. The entire chip is 60 square millimeters with a total of 56,340 components and 18,000 logic gates. It’s the most complex flexible CPU in the world — the runner-up has 12 times fewer components than the PlasticARM. It needs 21 milliwatts of power, which is high for the M0, but future versions could be efficient enough for IoT and wearable applications where it’s not feasible to charge devices frequently or keep them connected to power. Also, you can fold it in half while all this is going on.
ARM doesn’t have any specific plans for this chip — not only is it still a prototype, but it can also only run three programs that have been coded into the design. Future versions of the PlasticARM chip will have writable memory, allowing for full programmability.
The other issue preventing flexible SoCs from flooding the market is power efficiency. While the M0 core is a low-power design ideal for IoT technology, 99 percent of that 21 milliwatts is wasted. The overall chip is also larger than the silicon cousins. That doesn’t matter too much if it’s just going to be a flexible sheet of plastic, but the power drawbacks are a real hurdle. ARM research engineer James Myers told New Scientist what to expect as PlasticARM devices continue their evolution: “It won’t be fast, it won’t be energy efficient, but if I’m going to put it on a lettuce to track shelf life, that’s the idea.”
Nvidia Unveils ‘Grace’ Deep-Learning CPU for Supercomputing Applications
Nvidia is already capitalizing on its ARM acquisition with a massively powerful new CPU-plus-GPU combination that it claims will speed up the training of large machine-learning models by a factor of 10.
Star Citizen Developer Unveils New Roadmap, Cancels Squadron 42 Beta
Cloud Imperium Games has canceled the Squadron 42 beta that was supposed to debut before the end of 2020, with no current plan or timeline for launching it.
Framework Unveils Modular Laptop Coming This Summer 2021
Laptop startup Framework believes it's finally cracked the code on customizable mobile I/O.