HWiNFO beta 8.55 adds ASIC Quality percentage to show NVIDIA GPU factory efficiency

HWiNFO beta 8.55 adds an ASIC Quality percentage to display NVIDIA GPU factory efficiency based on SPEEDO values, helping buyers assess silicon quality.

HWiNFO beta 8.55 adds ASIC Quality percentage to show NVIDIA GPU factory efficiency

HWiNFO beta 8.55 build 6093 introduces a new metric that translates raw GPU factory data into a readable percentage. This change helps system builders and overclockers quickly assess the manufacturing quality of graphics cards without decoding complex voltage curves. Buyers can now see at a glance how a specific chip variant might perform under stock settings.

New beta metric translates immutable fuse data into readable silicon quality percentages

The update targets the SPEEDO values embedded in NVIDIA GPUs by factory fuses. These values define the default voltage and frequency characteristics for each chip. The new ASIC Quality feature calculates a percentage based on these immutable factory settings, providing a standardized way to compare silicon quality across different cards.

For the RTX 5090, the reported SPEEDO range spans from 2018.94 to 2284.87. One Founders Edition owner reported a SPEEDO value of 2176, which translates to an ASIC Quality reading of approximately 59.06%. This specific example illustrates how the software converts the raw fuse data into the new percentage format.

Higher SPEEDO values indicate better efficiency at stock settings, but they do not guarantee higher overclocking potential. Anecdotal evidence shows mixed results, with one enthusiast reporting an RTX 5090 with a SPEEDO value of 2207 that struggled to maintain 3200 MHz in games. Another user achieved competitive 3DMark results with a much lower SPEEDO value of 2076, proving that efficiency does not always equal peak performance.

This beta release gives users a clearer view of their hardware's baseline silicon quality. The metric focuses on factory efficiency rather than overclocking headroom, which helps manage expectations for performance tuning. The feature is currently available in the beta channel for those who want to monitor their GPU's specific characteristics.

Discussion

0 comments

Log in to join the thread with a thoughtful take, question, or correction.

Add to the discussion