Detailed Power Consumption and Overclocking
For power consumption measurements, we are still using the methodology that helped us discover the power draw issues of the RX 480 back in 2016.
Our testing involves the use of a National Instruments data acquisition device, which is intercepting the power being sent through the PCI Express bus (12V and 3.3V) as well as the ATX power connectors (12V) before they go to the graphics card and are directly measuring power draw.
On the game side, we are continuing to use Metro: Last Light for power measurement, given its history of being particularly taxing on GPUs in 4K.
In total power consumption, the RTX 2070 is nearly identical to the GTX 1080. The closest AMD competitor performance wise, the RX Vega 64, comes in at almost 100W more power than the two NVIDIA-based cards.
Taking a look at the power breakdown by connector, we can see the majority of this 170W is coming from the 8-pin PCI Express power connector with rest being provided through the 12V rail on the PCI Express slot. The 3.3V rail from the PCI Express slot is not used, as in the other Turing-based GPUs.
Using the latest beta version of EVGA's Precision X1 software, we put the EVGA RTX 2070 Black Edition through some quick overclocking testing.
In its stock configuration, the EVGA RTX 2070 Black Edition stabilizes at a clock speed around 1760 MHz, with temperatures staying just under 60C.
With a quick run through NVIDIA Scanner and an increase to the maximum available power target, we were able to achieve a clock speed increase of almost 200 MHz, to around 1950 MHz overall. Temperatures increased a few degrees but stayed right around the 60C mark.
It appears that as with the other Turing-based GPUs, the overclocking limit is caused by the inability to apply more voltage, and not temperature based.