This dashboard analyzes NVIDIA, AMD, Intel, Micron, SK hynix, Broadcom and Samsung inside a larger public-stock universe of 50 US and 50 non-US semiconductor names. It maps direct relationships, cash-flow direction, public evidence confidence, depth-5 supply-chain paths and an alpha score built around being early, under-discovered, undervalued and levered to AI earnings torque.
Within the seven requested anchor names, the model favors SK hynix and Micron because HBM/server-memory scarcity has direct AI cash-flow leverage and fewer “already fully discovered” penalties than NVIDIA/Broadcom. Intel scores as a turnaround option because its foundry option value is real but external foundry revenue is small and foundry losses are still large.
Average alpha by semiconductor layer and region. High-scoring layers tend to be HBM, advanced packaging, test/metrology, ASIC/connectivity and selected foundry/tool chokepoints.
Network uses all 100 public stocks as nodes and the highest-signal relationship edges for browser performance. Yellow nodes are the seven requested anchor names; blue nodes are US; purple nodes are non-US. Download the CSV for the full edge set.
Each row is source → target. “Cash-flow map” states who likely pays whom or whether the relationship is competitive/indirect. For many supplier/customer relationships, exact dollar amount is not separately disclosed.
Generated paths show how a company connects through up to five public-stock relationship steps, e.g., accelerator → foundry → equipment → materials → memory/OSAT. These are research maps, not proof of a single closed-loop invoice chain.