JiahuiChen99/Heterogeneous-Computing-Farm

This is an abandoned academic project (1406 days old, zero activity, no forks, only 2 stars). The description reads like a research thesis abstract rather than a production system. The core problem—orchestrating heterogeneous compute clusters—is already solved at scale by Kubernetes (with its federation and device plugin architecture), YARN (Hadoop), Mesos, and cloud-native orchestrators. The claimed differentiator (IoT device support, mixed OS/language environments) is either: (1) solved by existing platforms through plugins and protocol adapters, or (2) too specialized to have market value outside academic contexts. The project shows no evidence of: real adoption, active maintenance, documentation, or reproducibility from the GitHub metrics alone. With zero velocity and no community engagement, this represents a dead research artifact. Platform domination risk is high because AWS (ECS, EKS), Google (GKE, Anthos), Microsoft (AKS), and open-source Kubernetes already saturate this space comprehensively. The 'heterogeneous' angle is a known problem in cluster management; there is no defensible innovation here. A well-funded incumbent could trivially add IoT device support (edge compute is a hot area) or this could simply be absorbed as a reference architecture. Displacement is already complete—this project never competed in the first place.