CPUID's HWMonitor has been updated to version 1.65, introducing the ability to display thermal hotspot data for NVIDIA's RTX 50-series graphics cards. This new feature provides users with more granular information about their GPU's thermal performance, potentially aiding in the diagnosis of cooling-related issues.
The addition of hotspot sensor data appears as two new rows within the graphics card monitoring section of HWMonitor, situated below the standard GPU and memory temperature readings. This elevated temperature reading is expected, as hotspot temperatures are naturally higher than the average GPU temperature under load. High hotspot temperatures, especially when coupled with excessive fan speeds and reduced gaming performance, could indicate problems such as dried-out thermal paste or improper cooler contact.
To test the new functionality, an Asus TUF Gaming RTX 5070 Ti OC Edition was monitored while running Crimson Desert at 4K resolution with path tracing enabled during a UK heatwave. Under these demanding conditions, the average GPU temperature remained stable around 62-63°C. HWMonitor recorded a brief hotspot spike into the mid-80s, which then settled to a range of 75-80°C. These figures were deemed acceptable, consistent with previous benchmarks for the card, which also maintained quiet operation.
Concerns would arise if hotspot temperatures were to significantly exceed these readings, prompting a closer inspection of the cooler and its thermal interface materials. The developers of CapFrameX, another performance monitoring utility, have also announced plans to incorporate RTX 50-series hotspot temperature tracking into their software soon, offering users additional tools to keep tabs on their PC hardware.