Inspiration
In the United States alone, about 330 million smartphones sit plugged in, idle, and unutilized every night while we sleep. In our modern race to meet the growing demands of AI training, large scale simulations, and scientific research, we saw an opportunity to repurpose existing consumer hardware into a decentralized, sustainable alternative to resource-intensive data centers.
What it does
SpaceSync Cloud transforms iPhones into a distributed compute cloud. Our iOS app turns an idle iPhone into a remote worker node for computing JavaScript compute tasks remotely. All the user has to do is plug in the phone, run the app while they sleep, and they are contributing a small part of a massive super computer. They can donate this compute time for scientific sustainability research such as protein folding, or they can get paid fair market value for their Teraflops of compute power they contribute to our cloud.
How we built it
We built a Python FastAPI-based orchestrator that manages device registration, task queueing, and result aggregation. Our custom iOS app receives and executes JavaScript code securely in a sandboxed environment. Devices pull tasks, execute them on-device, and push back results over a REST API. SQLite is used for state management.
Challenges we ran into
- Building the execution layer for JavaScript on iOS using WebKit
- Balancing execution limits and battery/network constraints
- Designing a scalable architecture that supports any number of contributors
Accomplishments that we're proud of
- Successfully distributed and executed real workloads across multiple iPhones
- Demonstrated viable proof-of-concept for cloudless, crowd-powered computing
- Created a platform that promotes sustainability while unlocking real-world utility from idle hardware
What we learned
- Consumer devices are massively underutilized and capable of meaningful contributions
What's next for SpaceSync Cloud
- Expanding to Android would enable low-level system access, allowing us to build a true hypervisor-like environment and run more complex, AWS-like tasks, including containerization and deep learning workloads
Log in or sign up for Devpost to join the conversation.