Restored ROCm multi-GPU build flag for Windows in Ollama

{
  "event_id": "event_1cdb6a171d3aeddf",
  "topic_id": "gpu_supply_and_compute_market",
  "event_type": "Code Change",
  "event_time": "2026-06-01T17:44:22Z",
  "title": "Restored ROCm multi-GPU build flag for Windows in Ollama",
  "summary": "This pull request reintroduces a previously dropped ROCm build flag that enables multi-GPU support on Windows, specifically by fixing the amdhip64_*.dll version detection for the 'latest' selection, ensuring AMD GPU users can reliably use multiple GPUs for local AI inference.",
  "contribution": "Re-enables multi-GPU ROCm inference on Windows by restoring a critical build flag and fixing dynamic library version selection.",
  "impact": "Windows users with multiple AMD GPUs can now run Ollama without the multi-GPU feature silently breaking, which prevents crashes or fallback to slower single-GPU execution. Operators must continue to monitor for regressions in ROCm library detection across future Windows updates and AMD driver versions, as the 'latest' selection logic is sensitive to external changes.",
  "maturity": "Production",
  "confidence": 0,
  "importance_score": 0.7,
  "risk_flags": [
    "Driver Version Detection Fragility",
    "Windows Rocm Compatibility"
  ],
  "evidence_count": 1
}