MarbleHE/HECO

HECO (Homomorphic Encryption COmpiler) is a research-oriented project that addresses the significant complexity of writing FHE-compatible code by using an MLIR-based IR to automate noise management and SIMD vectorization. While the technical problem it solves is exceptionally difficult (earning it a baseline of defense), its quantitative signals are weak: 79 stars over 6 years with zero current velocity suggests it is a stale academic artifact rather than a living production tool. It faces stiff competition from better-funded and more active projects like Zama's 'Concrete' (which also uses MLIR), Google's 'FHE Transpiler', and Microsoft SEAL's own abstraction layers. The frontier risk is medium because while FHE is niche, it is a high-priority 'moonshot' for Google and Microsoft, who provide the underlying libraries HECO depends on. The displacement horizon is short because the state-of-the-art in FHE compilers has moved toward more integrated ecosystems that handle hardware acceleration (TPU/FPGA), which HECO lacks.