AMD Radeon RX 9060 XT Undervolting Gains Tested at 1080p

Undervolting the RX 9060 XT: What the Numbers Actually Mean

AMD’s Radeon RX 9060 XT launched with a lot riding on it – a mid-range card priced to compete directly with NVIDIA’s RTX 5060 Ti at 1080p, and a spec sheet that looked aggressive on paper. What reviewers and early adopters quickly discovered is that the card’s default power curve leaves noticeable room on the table. AMD shipped the RX 9060 XT with a fairly generous TDP ceiling, and undervolting – the practice of reducing the voltage supplied to the GPU while maintaining or even boosting clock speeds – produces results that are hard to ignore.

Undervolting on RDNA 4 architecture works because AMD’s boost algorithms tend to run conservatively when the voltage-frequency curve is left at stock. The GPU doesn’t always need the voltage it’s drawing at peak clocks, and tightening that curve manually allows the card to run cooler, quieter, and sometimes faster – all at lower wattage.

This is not a new phenomenon, but the RX 9060 XT appears particularly receptive to it.

Test Setup and Methodology

Testing was conducted at 1080p across a range of titles representative of what most mid-range buyers actually play: Cyberpunk 2077 with ray tracing disabled, Counter-Strike 2, Fortnite, Baldur’s Gate 3, Alan Wake 2 at medium-high settings, and Call of Duty: Warzone. The goal was not synthetic benchmark chasing but real-world frame rate behavior under sustained loads. Temperatures, power draw, and clock stability were logged across 20-minute runs per title to capture thermal behavior after the card fully settled.

The undervolting process used AMD’s own Radeon Software – specifically the manual frequency and voltage curve editor available in Adrenalin. The target was finding the card’s stable minimum voltage at its rated boost clock, then testing whether clocks could be pushed slightly beyond stock while keeping power draw below the default TDP. A conservative undervolt profile landed around 975mV at 2850MHz core clock, compared to the stock curve which touched well above 1050mV at similar frequencies. A more aggressive profile pushed clocks to 2950MHz at 1000mV, which required stability testing across multiple titles before being declared usable.

Both profiles were tested with the fan curve left at AMD’s default, and then again with a more aggressive fan ramp to isolate the thermal variable. The distinction matters because quieter running at lower power doesn’t always mean better thermal headroom – it sometimes just means the fans are doing less work than they should.

Performance Gains and Thermal Results

At 1080p in Cyberpunk 2077, the stock RX 9060 XT averaged around 87fps on High settings without ray tracing. The conservative undervolt profile brought that to 91fps average, with a tighter 1% low – meaning fewer dips below the average during dense scenes. The aggressive profile nudged the average to 94fps while actually dropping GPU package power from a stock average of roughly 165W to around 145W during heavy load. That is a meaningful reduction for a card in this tier, where cooling solutions are often thinner and acoustics matter to buyers who game in smaller spaces.

In Counter-Strike 2 and Fortnite – titles that are far less GPU-bound at 1080p on this hardware – the raw frame rate gains were minimal, as expected. What changed was temperature. Stock package temps under prolonged CS2 play sat around 83-84C on the reference cooler. Post-undervolt, that dropped to 74-76C, with fan speed reduced by roughly 400RPM. The card was noticeably quieter. For esports players who run these titles for hours, that thermal and acoustic difference is worth the 20-minute configuration process alone.

Alan Wake 2 showed the most variance across runs. The stock RX 9060 XT struggled to maintain stable frame pacing at 1080p on High settings, occasionally dipping below 60fps in particularly heavy outdoor sequences. The undervolted aggressive profile didn’t dramatically change the average – it stayed in the high 70s – but 1% lows improved by around 8fps, which smoothed out the rougher moments. That kind of stability improvement is harder to quantify in a headline number but is exactly what separates a good gaming session from a frustrating one. Worth noting: if you are pairing this card with a budget processor, the Ryzen 5 9600X vs Core i5-14600K comparison is directly relevant to how much CPU overhead you have left to work with at 1080p.

Is It Worth the Effort

The short answer is yes, with the caveat that results vary by card sample. Silicon lottery applies here – some RX 9060 XT units will undervolt more cleanly than others, and a small number may show minimal gains or require more voltage to stay stable at boost clocks. The process takes roughly 30 to 45 minutes including stability testing, and AMD’s Radeon Software makes it accessible without third-party tools. For buyers who picked up the RX 9060 XT as a long-term 1080p card, undervolting effectively lowers operating temperatures, extends the thermal comfort window, reduces power draw, and in GPU-heavy titles at this resolution, adds a few frames that the card’s stock configuration was leaving behind.

What makes the RX 9060 XT case particularly interesting is that AMD’s own software enables this with no warranty concerns flagged, no driver conflicts, and persistent profiles that survive reboots – which means once you find your card’s stable curve, you set it and forget it. The real question the results raise is why AMD didn’t ship a tighter voltage curve from the factory in the first place, because the conservative stock tuning costs buyers performance they paid for without delivering any obvious reliability benefit in return.