Advanced Tech: Silicon Carbon Batteries

Happen to run across this relatively new breakthrough in battery tech and spend time exploring it further yesterday. Some of you may have heard of Silicon Carbon Batteries being used in flagship phones with charger battery capacities. I think it could change how we store energy.

The key difference between SiC and normal lithium battery is in their anode composition. In a normal lithium-ion battery, lithium ions move in and out of graphite anode layers in a process called intercalation. In a silicon-carbon cell, tiny silicon particles sit inside a carbon structure and form alloys with lithium. This packs in far more Li ions than graphite. Meaning more energy density.

Silicon Carbide Battery

Theoretically, Graphite anode delivers around 372mAh/g. Silicon one can store up to 3,600 mAh/g(10x). But pure Si expand to 300-400% when battery is charged and contracts when discharged. This repeated size variation can cause the electrode to break down and not be useful in a battery. To confine this swelling, usually Silicon is coated with conductive carbon in SiC Batteries. There are also other secret sauces at play. So while we don’t get full 10x gains, we still achieve much higher capacity, boosting energy density from ~300 Wh/kg(normal) to ~450 Wh/kg(SiC).

These cells deliver higher energy for the same weight, making them ideal for both smartphones and EVs. They also charge faster, cutting down wait times significantly. With greater capacity per charge, devices need fewer charge cycles, which helps extend overall battery lifespan. It also means thinner, lighter batteries for the same performance.

In EVs, carmakers are testing them to push range to current max ranges or to cut pack weight. Stationary storage makers are looking at them for grid support, where extra energy and longer life helps cuts system costs. Chinese phone players are already using them, EV folks like Benz is considering them for their electric EVs. I think this tech will be mainstream in cars, phones, and home storage in the next 2-4yrs.

If you liked the post, Share it with your friends!

Advanced Tech: Neuton Edge AI

I got notified about Neuton AI, when Nordic Semiconductor acquired them last week. I have been playing around with the platform and the tools, and I am actually quite mind blown by what it can do.

Edge AI

Neuton is a no-code TinyML platform that builds ultra-compact neural-network models for small microcontrollers. You upload a CSV, press train & it grows the network, and you export it as a plain-C library that can be put on any microcontroller platform. The ridiculous part is the size of the generated model file is only around 5KB which is usually 10x lower than other ML platforms like TFlite, AutoKeras maintaining same or better accuracy. Small size also means inference time can be around 2ms. Faster inference implies lower battery usage too.

Edge AI

I really wanted to learn how they achieve this extreme size reduction. All the videos and content on the website are slightly vague. Their core tech is proprietary. Seems to have 2 US patents that cover architecture-free self-organisation and a parallel global search for weights and feature selection. What it means in practice is that the model stops growing the moment accuracy stalls, so there is nothing left to trim. They clearly mention they don’t use common training methods like back propagation or stochastic gradient descent. Each new neuron connects only to the most critical inputs or features rather than every possible input. This keeps the weight count low.

The platform keeps validation curves and lets you pick and download any smaller checkpoint if you prefer size to accuracy. It also has a signal processing engine to help with preprocessing of the data. Because the network connects only to the most informative inputs, it plays well with time series from IMUs, vibration sensors, ECG, and radar. I think this will let product teams skip a lot of manual DSPs and focus on features. Platform is also free to use, not sure if Nordic Semi fronting all the costs for the training runs.

For me personally, Neuton feels like a strong player for my future ML projects. Try it out yourself and see if it’s worth it for you.

If you liked the post, Share it with your friends!
1 4 5 6 7 8 10