The Z170-Claymore finished slightly down in 3DMark, but not by an amount that was noticeable in action scenes. A secondary examination with GPU-Z and CPU-Z showed that the top slot was continuously stuck in x8 mode, even though it has the required hardware to switch between x16-x0 and x8-x8 modes. The enhanced bandwidth of PCIe 3.0 should have been great enough to prevent even this small a difference, in theory.
The Z170-Claymore is in a dead heat with the top two competitors in memory bandwidth. All boards were tested using automatic settings, and the timing issues previously mentioned only occurred with “Manual” mode enabled.
Perhaps it’s time for us to re-evaluate our performance expectations with PCIe 3.0 x8 mode? The Z170-Claymore falls slightly behind the pack in both Battlefield 4 and Far Cry 3, though not by a great enough margin to impact actual play.
None of the tested boards really stand out in non-gaming applications, proving mostly that they’re all pushing the same CPU speed. Had there been a large difference, we’d expect something was wrong with the winner (or loser).
Power, Heat And Efficiency
The Z170-Claymore consumes less power than its rivals at idle, which is good for efficiency and expected in a board that doesn’t have many third-party controllers tacked-on. On the other hand, it consumes far more power under load. We opened CPU-Z to see why and were startled by the report (which turned out to be wrong). ECS informed us of the problem with CPU-Z, suggested we use a different program, and said that the board was operating as intended.
That still left me scratching my head over the power reading. I got out the voltmeter and found that even though CPU-Z was up to 0.50V off at maximum (reports too high) and minimum (reports too low) loads, the actual voltage topped out around 1.50V at stock settings. That’s around 250 mV higher than other boards.
The higher voltage also affected peak CPU temperature, allowing it to reach 80 °Celsius during extended load tests. At least the motherboard’s voltage regulator stayed cool.
A performance deficit of nearly 3% in games lead to an overall 0.7% drop in the average of all tests, which combines with a 16.6% higher-than-average power requirement to force a 14.8% loss of efficiency.