Clocks, Overclocking, Thermals, and Power Consumption
Here we will look at the physical and functional performance metrics for the new Z490 motherboards. This includes out of the box clocks, thermals, power consumption, and of course overclocking.
Out of the Box Clocks
First up is a frequency plot for the 10900K on the Z490 Aorus Xtreme, and we see the telltale TVB jumps to 5.3GHz. Still, under the Time spy stress test, we see the clock stick at 4.9GHz mostly as GIGABYTE has opened up the flood gates on the CPU as far as power and turbo limits due to their board design being able to handle way more than what this CPU can pull.
For power consumption tests, we use a wall meter to test the full system draw. The reason for this is it will represent what the entire system pulls versus our meter, which shows power draw on each PSU cable. The reason for this is that measuring the power draw from the EPS cable, for example, does not take into account VRM losses and, therefore, can show a much higher power draw for the CPU or other device due to an inefficient VRM design or loading range.
Idle power with the full system and TITAN RTX discrete GPU in place, we see the power usage sitting at 70W, which matches the Taichi and is six Watts shy of the Maximus Extreme. I simply think that while ASUS has several EC's that perform various specialized functions, it tends to cause the board to jump a few idle Watts over the other options.
Loading up the CPU with a full FPU load, we see that the Aolrus Xtreme jumps to the most power-hungry by a mere two Watts. This could be caused by various things from the overall out of the box boosting algorithm or adjacent presets and limits that GIGABYTE applies. This could also be an inherent slight loss of efficiency from the large VRM, but either way, it's only a couple Watts, so I'm not so sure it will be the end of the world to a user.
Testing synthetic GPU stress only is an excellent way to show the overhead the CPU adds, as the GPU does not pull more power after it reaches steady state. Here we see the Aorus Xtreme falls in line being a few Watts above the Taichi and a few more than the Supermicro. Then we have the Maximus Extreme, which pulls a lofty 21 Watts over the Aorus Xtreme. This may level out a bit with the new UEFI update as GIGABYTE boosted 3DMark performance according to the release notes.
Idle temps for the 10900K in each board are the same at 27C, with the Maximus Extreme variant dropping two degrees below the other three on our charts to a 25C result.
Loading each CPU up with a full burn-in FPU load, and we see the Aorus Xtreme nudge slightly over the Maximus extreme with an average of less than half od a degree or .4C to be precise. This is well within the margin for error on the three boards. The outlier here is the Sueprmicro Z490 board, which more closely follows intel guidelines to the number and therefore downclocks from its turbo boost speed after the 56-second turbo duration expires.
Here we see the Z490 Aorus Xtreme under our thermal imager with a full load on the CPU for over an hour. As you can see, the heat sink above and beside the CPU socket read 37 and 38C, respectively. The inductor line sits at around 42C, which is the hottest part measured and well within operating parameters with plenty of headroom.
Now time to check out the overclocking and thermals.
Overclocking with Comet Lake has been a lesson in limitations, both thermally and architecturally. As intel squeezes the 14nm node for all it has, the frequency is now under several boosting algorithms, which help enhance single or dual-core performance to the nth degree. However, it does not leave much on the table for overclocking.
The Z490 Aorus Xtreme was able to take our chip to the limit we found on previous boards. Our chip can do 5.3GHz if we push into ridiculous CPU temps well beyond what I would consider good for everyday usage. We settled at the same 5.2GHz clock at a 1.32VCore setting, which was immediately stable same as we have seen on other boards, but I did see that the GIGABYTE board seemed to overshoot slightly according to CPUZ. However, checking with a DMM, I see that the voltage is quite close in reality at 1.318Vcore.
Under load, it dips slightly into the 1.312V range, which was much better than I expected, and with tweaking of the LLC, you can probably tighten this up further. Overall as expected, we lose a few ticks on the stock out of the box single-core performance since TVB no longer kicks single-core boost into 5.3GHz range. However, the 5.2GHz all-core clock nets us roughly a 5% gain in performance. However, there is added heat and power draw, so keep that in mind as well.
Here we have a thermal image of the VRM area on the GIGABYTE Z490 Aorus Xtreme, and we see that the finned array cooler only reaches 42C on the side VRM and 41C on the top portion. This shows the potency of the GIGABYTE VRM cooling solution does a great job of being able to keep the power solution well tamed under the full fire of the 10900K overclocked load.
Last updated: Jun 3, 2020 at 04:05 pm CDT
- Page 1 [Introduction, Specifications, and Pricing]
- Page 2 [Motherboard Features and Marketing]
- Page 3 [Packaging and Accessories]
- Page 4 [GIGABYTE Z490 Aorus Xtreme Overview]
- Page 5 [PCB and Circuit Analysis]
- Page 6 [BIOS/UEFI and Software]
- Page 7 [Test System and Configuration]
- Page 8 [WPrime, SuperPi, Cinebench, and AIDA64]
- Page 9 [Handbrake, Blender, POV-Ray, CoronaRender, 7-Zip, and WebXPRT]
- Page 10 [Unigine and UL Benchmarks]
- Page 11 [System I/O Benchmarks]
- Page 12 [Clocks, Overclocking, Thermals, and Power Consumption]
- Page 13 [Final Thoughts]