Configuration:

A system with the following components was used for this review:

» Intel Core i7 920 quad core processor
» ThermoLab Baram CPU cooler
» ASUS P6X58D-E LGA 1366 Intel X58 motherboard
» Sapphire Vapor-X Radeon HD5870 1GB graphics card
» Seagate Barracuda XT 2TB SATA 6Gbps hard drive
» Panasonic slim optical drive
» Nesteq EECS 700 Watt power supply
» Windows 7 Professional 64-Bit operating system



With the Patriot Memory Viper II Sector 7 1800MHz 6GB DDR3 kit installed in the system above, I booted up and went straight in to the BIOS. I enabled the XMP profile, but found that the memory speed was set to 1600MHz. I then tried ASUS's DOCP (DRAM OverClocking Profile) setting, and was then presented with the memory set to roughly 1800MHz. Checking around the memory settings even further found that the voltage had been set correctly, and that the timings were set to 9-9-9-24, 2T. I relaxed the 24 to 27 and improved the 2T to 1T in order to match the specifications, and then rebooted in order for the settings to take effect.

After the system was allowed to boot in to Windows, CPUID's CPU-Z was loaded to see what that program had to say about the modules installed. The below left screen shot shows the "Memory" tab of CPU-Z, where we get confirmation that the memory is running at 900MHz (1800MHz DDR3), and that the timings are 9-9-9-27 1T. The below right screen shot shows the "SPD" tab where we see that these modules have two JEDEC profiles and the XMP-1800 profile programmed in.



Overclocking:

Previously, the fastest DDR3 triple channel kits I have used were all rated for 1600MHz, but with some effort I have overclocked them to 1800MHz and beyond. Since we were starting at 1800MHz as a base speed this time around, I was curious as to just how far they would go. Considering that the timings were already 9-9-9-27, I decided to only go as high as the kit would allow without relaxing things any more. I have taken some 1600MHz DDR3 close to 2000MHz, but it required relaxing the timings to CAS 10, and I have found that the performance at lower clock speeds and tighter timings is better.

By inching up the memory frequency in 5MHz increments I kept climbing until the system became unstable. When I hit that spot, I backed down to the previous good speed, and worked my way up in 1MHz increments. In the end, the maximum speed that could be achieved while still maintaining the 9-9-9-27 1T timings was 1920MHz. Not a tremendous increase over stock, but definitely a respectable total speed.