OCZ DDR3 PC3-17000 Gold XTE 4GB Kit - Benchmarking the OCZ DDR3-2133 Kit

Article Index
OCZ DDR3 PC3-17000 Gold XTE 4GB Kit
Test Setup and Overclocking
Benchmarking the OCZ DDR3-2133 Kit


Benchmarking - AIDA64:

Because there is a lot of system tweaking that goes along with clocking up memory to non-standard speeds, we haven't included a lot of "real world" benchmarks.  Often, in order to get the memory to run at speeds other than 1066MHz, 1333MHz, or 1600MHz, the motherboard bus and CPU must be clocked up in order to achieve these speeds.  As we've shown before the CPU plays a huge roll in benchmark performance and so it's not fair to compare WinRAR compression when the CPU is clocked up 400MHz faster.  Of course the "RAM" will look faster, but in reality the CPU is the one doing the work.

With that being said, the new Sandy Bridge Core i5/i7 platform has a native 1333MHz memory controller and many motherboards have options that allow you to change the memory multiplier to achieve fast memory speeds (up to 2133MHz on the P67A-UD3 from Gigabyte) without overclocking the CPU or bus.  This will allow us to show how simple memory speed improvements affect total system performance...or how little the affect performance.  The results are below.  First up is AIDA64 (formerly known as Everest).



When it comes to memory performance in AIDA64 - the higher number the better in terms of read/write/copy and lower latency numbers are better.  You can see that at stock DDR3-1600MHz, the Crucial kit holds its own even though the OCZ kit is running on CL7 timings.  As the memory speed increases, so does the performance and the latency even drops - although the CAS latency settings are raised.  The MHz performance speed offsets the great latency set in the BIOS.  It is impressive to see over 21,000MB/sec read speed through a dual-channel kit that is rated at 17,000MB/sec performance.


The other two tests are SuperPi and Cinebench.  We used SuperPi to calculate Pi to 1 Million decimal places using different memory speeds to see if memory speed affects programs that are largely CPU dependent.  The results pretty much speak for themselves.   Cinebench 11.5 was used in both CPU rendering as well as GPU / OpenGL modes to see how memory speed affects workstation performance.  I was surprised to see CPU scores stagnate while OpenGL score improved as the memory became faster.



SuperPi scores stayed basically the same regardless of memory speed.  They ranged from 11.045 seconds to 11.138 seconds when the CPU speed wasn't changed.  When we overclocked the CPU slightly to increase our memory speed to 2202MHz, we saw the largest improvement.  

Cinebench told a slightly different story.  OpenGL scores improved steadily as the memory speed increased, but CPU scores remained pretty much the same - only when overclocking the CPU did we see better performance on the OCZ kit.  The Crucial kit trailed behind in CPU performance but edged ahead slightly at stock DDR3-1600.



Final Thoughts:

OCZ Technology has done wonders for the RAM industry and has raised the bar over the past few years in terms of memory performance, aesthetics and pricing.  The Gold XTE series is their "mainstream high-end" kit that offers some pretty incredible speeds with looser timings in order to keep it affordable.  Running memory at speeds of up to 2.2GHz is pretty impressive and it stayed rock-solid at 2200MHz where it would fail at 2233 if we edged up the bus another single MHz.  They manage to push this memory to pretty incredible levels at 1.65v - which is higher than the JEDEC 1.5v standard - but is in line with many performance memory companies.

We did have issues trying to get this memory to run at anything faster than 1800MHz on a Core i7 920 X58 system as well as a Core i7 860 P55 system as these processor support memory controller configurations of up to 1066MHz - requiring a full 100% overclock of the CPUs memory controller.  Some people have got these systems to work, but we had to upgrade our test bench to a Sandy Bridge CPU that supports a native memory controller of up to 1333MHz - requiring less than a 100% overclock.  Also, as technology has improved we were able to easily hit these speeds without any issues at all.  We simply dialed in the BIOS setting and booted into Windows.


  • Large heat spreaders
  • DDR3-2133 at 1.65v
  • Runs at 7-7-7-21 1T at 1600MHz



  • Could not run DDR3-1866 at CL8 or DDR3-2133 at CL9




I'd like to thank OCZ for setting us up with some fast DDR3-2133 for this review.  I was impressed at getting over 20GB/sec of performance on a dual-channel kit and how smoothly everything worked when we used hardware that supports this speed of memory.

If you have any questions or comments, please post them at the "Comments" link below.