Sapphire P67 Pure Black Hydra Motherboard Review

Product: Sapphire P67 Pure Black Hydra Motherboard
Company: Sapphire
Author: James Prior
Editor: Charles Oliver
Date: June 6th, 2011



Overclocking the Core i7 2600K is simplicity itself, by design - the CPU multiplier is unlocked, and adjusting the platform base clock quickly renders the whole system unstable. Overclocking on Sandy Bridge is only possible by CPU multiplier adjustment. With Intel's second generation Core i-series, multi core turbo modes are available. Disabling turbo modes and increasing the non-turbo ratio had no effect on final CPU clockspeed, so for our overclocking endeavors we simply set all core ratio limits the same.

Sapphire TRIXX MB

Our final overclock was with a ratio of 46, giving 4600MHz - the math is nice and simple with a 100MHz base clock. This was achieved at 1.47v vCore, with 75% loadline calibration setting. We were able to POST and boot into Windows all the way up to x50, but stability was never achieved and random BSOD's were order of the day, even at 1.65v vCore, with hard locks and BSOD's after a couple of hours stability load testing. 4.6GHz is quite a lot higher than the overclocks on our test comparison systems, both of which achieved 4GHz.


Below you can see the power draw of the test system, under idle, media playback and artificial load conditions. Measurements were recorded using a Kill-a-watt meter to record the total system power from the wall. Idle conditions were measured at Windows 7 aero desktop with Balanced power profile and ASUS EPU enabled in Auto setting. Media playback results were taken during playback of AVC 1080p encoded video, using Cyberlink PowerDVD 10 Mk II and AMD Stream/APP disabled. Load power test results were taken during a 15 minute Prime95 in-place FFT torture test. Highest observed readings (Power drawn, in Watts) during the test period recorded, lower values are better.


Below you can see the CPU core temperatures of the Core i7 2600K when used with our HAF X and Noctua NH-C14 CPU cooler. Temperatures were recorded using CPUID's HWMonitor for all cores, and the highest readings (core temperature, in degC) during the test period recorded, lower values are better.

Memory Bandwidth

Results of the Read, Write and Copy performance benchmark from AIDA64. Higher speeds are better. Stock clocked AMD systems are running DDR3-1333, and DDR3-1600 for overclocked. For Intel, stock clocks are DDR3-1333 and DDR3-2000 for the overclocked Core i7 920 (triple channel 3x2GB), DDR3-1333 for the Core i7 2600K (dual channel 2x2GB). AMD systems were configured in unganged mode, running 3x2GB configuration.


In operation, Noctua's NH-C14 cooler is very quiet; the two 140mm fans are quieter than the system fans in our CoolerMaster HAF X case (equipped with optional additional 200mm top exhaust fan). Noctua's NH-C14 offers four modes of operation, based on fan installation - none, bottom only, top only, and both. When in bottom only mode, it is the same height as a full height add-in board - great for smaller but not half-height form factor cases. Removing the lower fan gives more clearance for board components, depending on how you want to use the additional airflow the down flow design - aim it over your RAM or your power circuitry, perhaps your chipset, depending on design and setup. Anecdotally, we found the NH-C14 cooler is capable of keeping the Core i7 2600K under 75C when overclocked to 4GHz without requiring the fans to be operational. Above 1.4v vCore we found the fans to be necessary, which dropped temperatures dramatically.