Z87 Stinger Firmware
The Z87 Stinger is a formidable competitor in multiplier-based overclocking. But if you're doing any tuning with the BCLK, then beware. A “d6” error code had this sample locked up for hours after setting 104 MHz with integrated graphics disabled.
It turns out that whenever the board experiences overclocking issues on its PCIe bus, it reverts to integrated graphics. If integrated graphics are disabled, it stops. And the CLR_CMOS button does not reset the base clock nor re-enable on-board graphics. Pulling the battery with the power disconnected and allowing it to drain down for several hours didn’t solve the issue, nor did changing the CPU or graphics card. Even EVGA’s own GeForce GTX 760 in our lab didn’t solve this show-stopper.
Forcing configuration errors is my last-ditch approach to sneaking into a board's firmware without replacing its CMOS ROM. In theory, a platform should (and usually does) jump to default settings when major hardware changes take place. The problem is that all of my LGA 1150-based processors are the same model!
I tried booting without memory and then booting with memory, but the UEFI remained stuck at 104 MHz with on-board graphics disabled and a PCIe error to the graphics card. I even tried removing the CPU, starting without the CPU, and then re-installing the CPU. Many of these changes would cause the board to temporarily use default settings, but it would then detect the hardware and reboot with the previous problematic settings before POSTing.
Thank goodness for bad memory, though. After pondering the type of error that would force the Z87 Stinger to remain in default mode until I could enter the UEFI, I remembered that some very old DDR3 modules aren’t properly programmed for newer memory controllers. I reached into my scrap pile, pulled out some discarded memory, and installed it. Viola, the error was detected, the board couldn’t figure out how to re-configure it at previous settings, and I was dropped into the firmware for manual configuration.
Remember that the diagnostics process started with many attempts to CLR_CMOS? Upon entering the UEFI, I found that the 104 MHz setting was still locked and loaded for application upon the next boot. I also found that on-board graphics would still be disabled. Resetting to defaults from inside the firmware got me back up and running.
The Z87 Stinger reached our expected 4.6 GHz at 1.25 V using a motherboard setting of 1.245 V, but couldn’t configure our memory to its XMP-3000 settings. It instead chose a 30x memory multiplier that’s not supported by any LGA 1150 processor. The board was, however, able to reach DDR3-3025 by choosing its DDR3-2933 setting and then increasing BCLK to 103 MHz.
Primary, secondary, and tertiary timings are all configurable from the Z87 Stinger’s memory menu. A 1.635 V setting produced a detected 1.65 volts.
I’ve never used memory training algorithms to search for higher DRAM clocks, but welcome the opinions of advanced memory tweakers in our comments below. The Z87 Stinger provides a lengthy selection of these settings.