Test System

Initial Setup

I installed the sticks and then the Noctua NH-C14 we normally use for all testing. Due to the cooler's size, it serves as a good example of how DIMMs and heatsinks may interfere with each other, however, I noticed no problems at all with these Patriot PX538G2000ELK sticks. As you can see in the second and third images above, there's just enough room between the Noctua's lower fan and the top of the Patriot PX538G2000ELK DIMMs, but it's not a lot of room, with just barely over a millimeter of space is left between the two. The angled top of the DIMMs will help direct airflow from the Noctua's fan over to the other DIMM in the pair, and I cannot help but think that Patriot, having over 25 years of experience, did this part of the DIMM design quite intentionally. It's not a big thing, but all the little things together do add up to a very capable cooling solution that's guaranteed to ensure these modules do not overheat.


The modules measure in at 41.87 mm in height; just a wee bit taller than the G.Skill DIMMs I reviewed several weeks ago. This also means that the PX538G2000ELK sticks are significantly taller than the Samsung LP DIMMs we looked at last week, so they may interfere with some aftermarket cooling options, but not all, as evidenced by my installation with the Noctua cooler.


Of course, because these PX538G2000ELK modules are XMP-enabled, there are two profiles for these DIMMs; first, the JEDEC profile, shown in the three images above, and secondly, the XMP profile which is shown in the three images below. It's very nice to see that these Patriot modules use a 1066 MHz JEDEC profile, which will help ensure that they boot in nearly any system that they might be installed into. It's also worth noting that due to the testing platform used, I did have to manually adjust things to get proper XMP operation.


The XMP profile, shown above, is actually not 100% compatible with our Intel X79 Express-based test platform, so I did have to boost the BCLK to 125 MHz in order to get the default 2000 MHz to work. Without doing so, these sticks will default down to 1866 Mhz with the timings and voltage of the 2000 MHz profile, which, if I used that for testing purposes, would not convey the actual performance offered by these sticks as that is a platform limitation, not a limitation imposed by the DIMMs themselves. Of course, this does highlight why I use the Intel X79 Express platform for memory testing, as its flexibility is fantastic compared with the seemingly "locked-tight" options that the SandyBridge-based platforms offer in regards to "tweakability".

Now, I do have to mention here, that with a default 100 MHz BCLK, just enabling the XMP v1.2 profile resulted in the system failing to complete POST testing, even though the DIMMs were set to "downclock" to 1866 MHz. XMP v1.2 profiles do not contain information for VCCSA voltages, which is critical for proper operation of these DIMMs on the Intel X79 Express platform, even with just two sticks installed. However, knowing the platform, it wasn't hard to figure out exactly why this occured, so just wait and let the board reset itself to default settings in order to boot as it's supposed to on a failed boot, and then I went into the BIOS and boosted the VCCSA voltage to 1.05 V, which is a common voltage for DIMMs of this speed with XMP v1.3 profiles.