|The Test Rig:
Intel E8400 @ 3.00GHz
Above we have the CPU-Z screenshots with the E8400 at its stock settings and the memory voltage set to 2.1V with the appropriate timings of 5-5-5-18 2T. Upon first boot it automatically comes up at 800MHz with the voltage at 1.8V. Having had issues with memory in the past. I have kind of grown into the habit of setting the memory to it’s specifications manually as soon as I boot the build up. This will allow you to know that they should perform as rated. Believe it or not, I have seen several people build a rig and 6 months to a year later they realize that their memory has been running at a lower clock than what they was rated for. When they set it appropriately, they was like “WOW, it’s like a new computer”.
On to the testing process we go. For this review I will be showing a couple of other memory options. Unfortunately, both comparison modules are of the DDR2 800MHz variety and are only 2GB. The first is OCZ’s SLI-Ready memory and the second is Corsair’s XMS2 DDR2 800MHz. We will be using Sandra Lite V 2009.1.15.42, Cinebench R10, 3D Mark Vantage, Super-Pi Mod1.5, Team Fortress 2, and Call Of Duty 4 to test each of the modules at different frequencies. The E8400 is rated at 1333MHz FSB @ 3.0 GHz out of the box. Our sample has been tested to over 4.2 GHz on air with previous configurations. To keep things on a level playing filed. I will simply back the stock multiplier down to 8X from 9X. I will then manually set the FSB it to 375×8 which will put me back at the 3.0GHz mark. I will then adjust the memory accordingly from 800MHz up until I find its maximum stable frequency. Each of the tests will be ran three times and then the average of those three runs will be recorded.
We selected Sandra Lite as our choice of software to measure the memory bandwidth, which is being measured in GB/s. Both of the DDR2 800Mhz modules started developing instability issues at frequencies above 1072MHz. The Reapers where tested at 1066MHz, and then their maximum (1090MHz) frequency. We can see by the above results, that all three sets of modules came out pretty close to each other. I was a bit surprised by the amount of the jump I seen on the Reapers going from 1066MHz to 1090MHz. Now, keep in mind, that a higher transfer rate does not mean that your system is going to achieve a higher overall performance level.
Cinebench is designed to give you a real-world performance test. It is based on Maxon’s animation software, Cinema 4D, which is used by several large studios to create 3D content that for movies. Cinebench is capable of testing up to 16 CPUs or CPU cores. It is available for Windows (32-bit and 64-Bit) and Macintosh (PPC and Intel-based) systems. By looking at the above results, we can tell that it is very Dependant on the CPU. Although we do see a slight difference between the results, there’s really not enough difference to say one out done the other.
3D Mark Vantage:
We used 3D Mark Vantage next to test the systems overall performance. As we can see by the above results, the OCZ Reaper HPC modules did carry the best overall scores. So far the difference between all three modules has been minimal, which could still leave a consumer in the position of not knowing which way they would want to go. The biggest difference between the three modules has been the 2GB vs 4GB. The Reapers being the only 4GB option. Vista has shown to run its best with 4GBs or more.
Super Pi Mod1.5:
I then made multiple runs on Super Pi. I used the 1M calculations and repeated the tests 3 times recording the average between the three runs. The Reapers did achieve the lowest score of the three modules. Unfortunately it took a dive for the worst when we bumped it up to 1090MHz and resulted in the worst time. This could be the result of the slightly more relaxed timings.
Team Fortress 2:
For Team Fortress 2, I tossed in a full round demo and used the timedemo command to pull the average FPS achieved through out the round. I ran each set of memory through three passes and averaged the results. We see the OCZ Reapers trailing by a few FPS here at the higher frequencies, but leading at the lower frequencies.
Call Of Duty 4:
Last but certainly not least, we fire up Call Of Duty 4 and utilize the timedemo command again. Here we see both of the OCZ modules showing slightly better FPS overall when compared to the Corsair XMS2 modules. The OCZ Reapers, just barely nudge out the lead against the OCZ SLI-Ready modules.
Unfortunately, my overclocking results on the OCZ Reapers wasn’t to pleasing. I was able to get the modules to run a maximum frequency of 1100 MHz with the CPU multiplier backed all the way down to 6x, but once I started bumping the multiplier back up I had to bump the modules back to 1090MHz to get the modules stable, which is only a 2.25 % increase. Playing with the voltage and timings didn’t really offer much of a gain. I actually tested two different sets of the OCZ Reaper HPC’s and had the same outcome with both.
Being satisfied with the stock performance and looks of the modules. I picked up a second pair and gave them a new home in my Green Flame mod that debuted at NVISION 08