We’re all sitting on a computational stockpile that is largely untapped

Nosana
4 min readFeb 23, 2023

Did you know that manufacturing electronic devices such as smartphones is an insanely energy-intensive process, accounting for 70–85% of the lifetime carbon footprint of the phone itself? Still, we’re seeing 150 million smartphones discarded every year in the United States alone, amounting to one phone discarded per person every two years. And those phones are not properly discarded or recycled but simply forgotten in drawers and storage units, creating a computational stockpile that is largely untapped. So we’re all sitting on this computational goldmine while damaging the environment with our behavior. Not ideal, right?

Processor graveyard
Most of these smartphones are just sitting in junk drawers and storage units because most people don’t know about recycling programs or how to use them for other things. To make matters worse, many phones aren’t designed with recyclability in mind, meaning that even when they do find their way into recycling centers, they may still end up being incinerated or landfilled due to their complex internals.

This wasted potential is especially troubling now that smartphones have more powerful processors, better ways to connect to other devices, and reliable power sources. These devices could be used for more than just making calls. They could be used to give remote access to medical services, help monitor the environment, or, in our case: decentralized computations for CI/CD jobs.

New life for old processors

Nosana is based on the idea that employing existing digital devices as “computational nodes” can solve a variety of problems. And the research backing these claims is popping up more and more often. Take this research project by scientists from the University of California, Berkeley, for example: https://dl.acm.org/doi/pdf/10.1145/3575693.3575710.

The researchers suggest that repurposing phones could help with the growing global compute capacity and reduce carbon emissions. They go on to verify that mobile devices can indeed be repurposed as “junkyard computers” to do all kinds of computing tasks. The report even shows how decade-old smartphones can meet the performance demands of modern cloud microservices and be combined to perform increasingly complex tasks. Sounds familiar, right?

The research introduced a new carbon-aware performance metric known as Computational Carbon Intensity (CCI). A metric measuring the lifetime carbon impact of a device versus the lifetime useful compute it performs. They used CCI to quantify the value of extending the service lifetime of computational devices. What’s interesting is that the researchers applied CCI to old servers, laptops, and smartphones and found that used smartphones repurposed as general-purpose compute nodes offer the best potential for carbon impact. With approximately 60–70% of decommissioned smartphones neither thrown out nor recycled, this is a remarkably large potential source for repurposing.

So we like to take it even one step further. Why focus on decommissioned smartphones alone? In the world we live in now, many devices, like smartphones, laptops, and tablets, can do complex computing tasks that used to be possible only on high-end computers. However, most of the time, these devices are not fully utilized, leaving their spare computational power untapped. This is where the idea of “distributed computing” comes into play, which involves leveraging the unused computational resources of a larger network of devices to perform complex computational tasks. There are many potential benefits to using spare compute for computing. For one, it allows for more efficient and cost-effective processing of large amounts of data.

Pooling resources

We believe we have found the way to achieve this goal: through the use of blockchain technology. The technology allows for the secure and transparent recording of transactions on a decentralized network. We leverage the blockchain to create a decentralized computing network, where devices can contribute their spare compute power to a shared pool of resources.

Distributed computing utilizes the collective power of multiple devices to complete complex tasks. By leveraging the combined processing power of individual devices, distributed computing using blockchain technology can reduce the need for large server farms, resulting in lower costs and improved performance. The shared pool of computational resources is securely and transparently managed, without any single entity having control over the network, thereby making the network resistant to any attempts at centralization or control by any single party. In the meantime, users receive compensation for contributing their spare compute power based on supply and demand in a fair market.

Overall, turning digital products into computational nodes is an exciting development that has the potential to revolutionize how we use technology in the future. As more companies invest in this concept and develop solutions to address the challenges associated with it, we may soon see it become commonplace across a range of different industries.

Thanks for reading!

Unlisted

--

--