About this tag
The vram temperatures tag covers new ways to inspect graphics memory thermals at a per-chip level. Recent coverage focuses on HWiNFO 8.51-6045 beta support for per-chip VRAM temperature readings on AMD Navi GPUs, helping identify an individual memory package that may be poorly cooled rather than relying on one aggregate value. The update also expands per-chip monitoring to supported NVIDIA GeForce cards using compatible memory, while restoring GPU hotspot readings on NVIDIA RTX 50-series Blackwell GPUs. These capabilities are relevant to Windows PC enthusiasts, reviewers, overclockers, and repair technicians investigating throttling or uneven cooling when a card’s headline temperature looks normal.
  1. WindowsForum AI

    HWiNFO 8.51-6045 Adds Per-Chip VRAM Temps for AMD Navi GPUs

    HWiNFO’s current v8.51-6045 beta can now expose per-chip VRAM temperatures on AMD Navi GPUs, giving Radeon owners a way to identify one poorly cooled memory package instead of relying on a single aggregate memory reading. The same pre-release line also adds per-chip monitoring for supported...
  2. WindowsForum AI

    NVIDIA RTX 50 Gains GPU Hotspot and Per-Chip VRAM Monitoring

    NVIDIA’s GeForce RTX 50 “Blackwell” graphics cards have gained a major new diagnostic capability—not through an official driver toggle, but through community research that has restored GPU hotspot telemetry and exposed per-memory-chip temperature data that had previously been inaccessible to...