Knowing full well that the Ivy Bridge processors tend to run hotter than their Sandy Bridge counterparts, we’ll dive in to the overclocking section with the hopes of getting our customary 1 GHz overclock. Well, OK more like a 1.1 GHz overclock, because I’m actually shooting for 4.6 Ghz. The stock speed of the i7 3770K is 3.5 GHz; we’ll see if we can tame the temperatures well enough to allow reaching our goal.

Here is the rundown on the components used for the testbed:

• EVGA Z77 FTW BIOS Version 103 (Alpha version, yet to be released)
• G.SKILL Trident X (2 x 8GB) DDR3 2400 F3-2400C10D-16GTX
• Kingston 3KSSD 240 Gb
• Intel i7 3770K Processor
• EVGA Superclock CPU Cooler
• Corsair HX1050 PSU

Once I started the overclocking, I ran into a few memory related issues with the BIOS. EVGA, in their customary fashion, was quick to respond and had a new BIOS in my hand within a few hours. The alpha release of BIOS 103 fixed the issues I was having, and should be an official release very soon. I went ahead and jumped right to 4.6 GHz, believing that to be the maximum overclock that allowed temperatures to remain at acceptable levels under air cooling. My hunch was correct and at 100% CPU load the temperatures maxed out just over 80C.

Getting to the 4.6 GHz overclock was easy to accomplish with just a bump in CPU volts to 1.250, and changing the CPU multiplier to 46X. The memory ran solid as a rock at their rated 2400 MHz speed, and remember, these are 8 Gb sticks!.

Below are some screenshots of EVGA’s E-LEET software showing the Overclocked speed, memory settings, and the voltages. E-LEET can also be used to Overclock from the desktop, including CPU multiplier adjustments and all the voltages. The monitoring tab gives you real time information on fan speeds, temperatures, and voltage readings.

Once set to 4.6 GHz and the memory running at 2400 MHz,  I ran wPrime 32M and 1024M . A sub 6 second 32M run is pretty impressive! The run at 1024M was an impressive time as well, and would definitely point out any instabilities if present.

A sub 8 second result from SuperPI 1M was produced next; followed by a SuperPI 32M run at almost exactly 7 minutes. The 32M run is another good indication of the system stability. I’ve seen cases where Prime95 and LinX stress test utilities pass with flying colors, but the 32M run of SuperPI will still fail to complete.

The last stability test was using AIDA64 for a quick 5 minute run. As you can see, one of the cores was maxing out around 82C, these Ivy Bridge CPUs need WATER!!