Akash Launches Homenode for Decentralized Compute
Akash has launched the first phase of its Homenode project in early access. The platform enables individuals to contribute compute power from devices equipped with high-end GPUs, such as the RTX 4090 and 5090. The initiative aims to expand compute access for AI developers and enterprises by leveraging a decentralized network of hardware.
- The initial "Phase 1" of the Homenode open beta is focused on benchmarking performance using high-end NVIDIA GPUs, specifically the RTX 4090 and the upcoming 50-series. - This initiative is a core component of Akash's broader "StarCluster" project, which aims to build a globally distributed network for AI compute. - The Homenode architecture is designed to simplify the process for non-technical users by having the Akash Core Team at Overclock Labs manage the centralized "control nodes," while individuals run the "worker nodes." - According to Cheng Wang, COO & CFO at Overclock Labs, the goal is to "harness every available teraflop," and by starting with the most powerful consumer GPUs, they can establish high-performance standards before expanding support to other hardware. - The project aims to tap into underutilized hardware in homes to help alleviate the energy consumption constraints faced by traditional, centralized data centers. - Akash's broader network has seen significant growth, with deployments increasing by 466% to 3.1 million in 2025 and network spending more than doubling to $3.2 million. - Homenode is part of a larger 2025 roadmap for Akash to transition from a resource network to a services network, with a key priority of improving the client experience to remove barriers to adoption. - The project is entering a competitive landscape of decentralized GPU platforms, including networks like io.net, Render, and Aethir, all aiming to provide more cost-effective compute for AI developers.
