Test System

Initial Setup

Initial setup proved very easy, with everything working right out of the box, and no obvious problems. However, as you can see in the screenshot above, the CPU speed is listed as 3492 MHz, however, once we cleared the CMOS, we found that the CPU was stuck to a 38 multiplier. Upon careful investigation we discovered that the Intel Turbo Boost modes were not functioning correctly, and there was no clock scaling according to how many cores were loaded, leading to a 3800 MHz default clock on all four cores, rather than the 3500 MHz that is standard for our CPU. This will affect the benchmark scoring, of course, but as that's the default according to the BIOS, that's what we went with. As you can easily select the CPU multiplier, it's definitely possible to set the stock 3500 MHz clocks, but single- and dual-threaded workloads will see a performance drop in comparison to other boards.


The multiplier issue wasn't the only problem we encountered. The images above show the backside of the board and two of the pushpins from a stock cooler. Those surface mounted components you see there in the images are inside the square made by the mounting holes, and due to this, we could not use our Noctua NH-C14 cooler, nor could we use the factory provided mounting for our Corsair H70 cooler, as the backplates from both coolers would have rested on top of these surface mounted parts, leading to uneven cooler contact. We are fairly sure that this will be the case with nearly every cooler on the market that uses a backplate, and while this might be an issue for some users, for us it was completely expected given the incredibly small size of the Z68-ITX WiFi. There just simply isn't any room, whether it's the front of the socket, or the rear, for anything much more than a stock cooler.


So, with that said, for the first time, we had to do the majority of our testing using a stock Intel cooler! Although Sandy Bridge CPUs literally sip the power, at 4.5 GHz consuming up to 125 watts, they can still get quite hot. We noticed 90°C load temps reported by the CPU's internal DTS sensor with 3.8 GHz clocks (using the Intel cooler). Anything beyond that, not even with voltage increases, would lead to thermal throttling.

If this was any other board, it would be reason for concern, but as you can see in the image above, the board is really really tiny, dwarfed by the reference Sapphire HD 6950 2 GB we use in our test bench.

PWM Power Consumption

Since one of our first tasks was to truly verify system stability, while doing so, we measure CPU power consumption. We isolate the power coming through the 8-pin ATX connector using an in-line meter that provides voltage and current readings, as well as total wattage passed through it. While this may not prove to isolate the CPU power draw in all instances, it does serve as a good indicator of board efficiency and effective VRM design.


For idle power consumption, the ZOTAC Z68-ITX WiFi ended up in the middle of the pack, with our meter reporting 10W consumed. However, this changed quite drastically under load, with the Z68-ITX WiFi giving us the best numbers we have seen yet, beating out the best results we got from the ASUS P8P67 PRO by 2 W. Very impressive!