Sustainable, Autonomous, and Decentralized Infrastructure for the AI Era.
Grand Finalist — Binance Labs Incubation Program
Solar DePIN is a hardware-and-software ecosystem designed to bridge the gap between renewable energy systems and decentralized high-performance computing (DePIN/AI). By deploying custom, solar-optimized GPU nodes directly at commercial and residential solar power installations, we enable 100% green, zero-carbon, and highly cost-efficient cloud computing for the global AI/ML market.
We don’t just have a concept — we run a live, green technical setup:
- Energy Generation: 30kW Solar Power Station (PV Array) with a backup Uninterruptible Power Supply (UPS).
- Inverter System: Huawei SUN2000 with remote telemetry support.
- Compute Node: High-Performance Server running AMD Ryzen 7 CPUs and NVIDIA RTX 3090 / 4090 GPUs.
- Containerization & OS: Docker-based environment, Linux (Ubuntu LTS).
- Monitoring & DevOps: Prometheus + Grafana + Custom Watchdog script for 100% uptime.
We go beyond standard servers by developing our own specialized physical hardware:
- The Prototype: We are in the final stages of assembling and testing our first custom physical Solar DePIN Node prototype, set to be unveiled very soon.
- Solar-Inverter Link: An integrated microcontroller that natively communicates with inverter protocols without requiring complex third-party software setups.
- Flexible Economic Architecture: Our economic model is built to maximize profitability for solar owners without regulatory risks, utilizing secure node runner rewards. Concept: We are building a green, eco-friendly AI compute cluster consisting of 10 micro-nodes. Each node is a standalone, autonomous mini-server based on Intel N100 (or similar x86 Intel CPU/NPU with 16GB+ RAM). Combined, these 10 nodes form one distributed super-server powered by our 30kW solar energy architecture. Tasks for the First Physical Node (Proof of Concept):
- OS & Environment Setup: Install Ubuntu Server 24.04 LTS. Set up Docker + Docker Compose to run isolated AI workloads.
- AI Runtime (Model Inference): Deploy Ollama or vLLM inside Docker to handle lightweight LLM models (e.g., Llama 3 8B, Phi-3, Qwen) utilizing Intel iGPU / OpenVINO acceleration. Expose an API endpoint so this node can seamlessly connect to our central cluster/orchestrator.
- Inverter Integration (Modbus RTU): Connect the node directly to our solar inverter via a USB-to-RS485 adapter using the Modbus RTU protocol. Deploy a lightweight microservice (Python / Go) to poll live solar generation data and battery SoC (State of Charge). Dynamic Load Balancing: Implement logic to adjust AI workloads based on solar power yield — when solar power is peak, maximize AI task processing; when generation drops, throttle tasks or switch to idle/standby mode.
- Monitoring Stack: Install Prometheus Node Exporter to push hardware metrics (CPU load, RAM usage, core temps, Modbus connectivity status) to our primary server's Grafana dashboard.
- Binance Labs: Achieved Grand Finalist status in the incubation program.
- Proof of Concept: Successfully launched our live 30kW solar testbed with high-end RTX 3090 GPUs.
- Stability Framework: Integrated automated monitoring tools (Watchdog, Prometheus, Grafana).
- Hardware: Assembling and testing our first proprietary physical Solar DePIN Node prototype.
- Akash Network: Defending our Ecosystem Proposal to deploy solar-powered Akash nodes via our hub.
- Solar Gateway: Releasing the first stable software version to pull real-world metrics from Huawei inverters.
- Closed Beta: Rolling out the software to a select group of early solar home-system owners.
- Dynamic Scaling: Achieving fully automated, sun-cycle-dependent AI workload scaling in beta.
- Hardware Partnerships: Collaborating with components distributors to lower node production costs.
- Mass Production: Launching the first commercial batch of plug-and-play Solar DePIN nodes.
- Multi-Inverter Support: Adding compatibility for Fronius, Sungrow, Victron, and other top brands.
- Economic Launch: Finalizing and deploying the reward mechanism to incentivize green node operators.
We are actively seeking strategic cooperation in the following areas:
- Grants & Funding: To purchase components, scale testing, and mass-produce our physical GPU nodes.
- Hardware Partnerships: Strategic collaborations with solar inverter/panel manufacturers and hardware integrators.
- Ecosystem Collaborations: Integrating our green computing power directly into decentralized cloud marketplaces (e.g., Akash Network).