Overclocking
When it comes right down to it, overclocking is all that matters to a lot of people when it comes to memory. This is understandable, since the faster you can clock your memory, the more performance you can squeeze out of it, in theory. If you have to go to CAS Latency 2.5 to get an extra 10 MHZ out of your memory, it probably isn't worth it, as the increased MHZ isn't worth taking the timing penalty. XtremeDDR DDR400 True comes programmed with very modest SPD timings of CAS 2.5 and 2T command. The SPD hasn't been programmed with faster timings because, according to XtremeDDR, great stability and good performance are better then bad stability and great performance. The logic is there, but is it truly logic? Is it possible that the modules can't actually reach their intended speeds at faster timings? Let's find out.
Overclocking Test System
Abit KX733 Mainboard BIOS Rev. 1009
MSI KT3 Ultra - **Would not boot**
Unlocked AthlonMP 1800+
1x XtremeDDR DDR400 True
ThermalRight SK6 Heatsink with 60mm Delta Fan
Sandra 2002
Before I go on with the review, I would like to explain something. In previous testing XtremeDDR PC2700 modules would not work with the MSI KT3 Ultra. The same holds true for their DDR400 True modules, as the system locks up during the Windows XP boot screen when the modules are being used. XtremeDDR's explanation is that the memory uses an aggressive SPD, but I am not quite sure if that is the only culprit.
The memory booted up and worked fine in our Abit KX7333, so we will be using it for testing. In order to get the best possible overclocking results I used a 5:2:1 divider and kept the processor at or below 1500 MHZ via a multiplier at all times.
133 MHZ CAS 2 1T COMMAND-Sandra 150 Run Loop Test - PASS
As expected, the memory easily passed the 150 Run Sandra Memory Burn-in. After the burn-in the system was responsive and I was able to complete tasks such as opening windows and surfing the internet. The memory was running at 2.6 volts.
166 MHZ CAS 2 1T COMMAND-Sandra 150 Run Loop Test - PASS
We jumped from 133 MHZ to 166 because anything in between would have been a waste of time. The memory ran fine at the fastest timings available, with an "Ultra" setting. Previous incarnations of XtremeDDR have been able to run at 166 MHZ under the fastest available timings, and this was no exception. The system was responsive under these settings before and after the benchmark. The memory was running at 2.6 volts once again.
176 MHZ CAS 2 1T COMMAND-Sandra 150 Run Loop Test - PASS
Bumping the processor down to a 10x multiplier to keep it in line, we were able to run the memory just fine at 176 MHZ at the fastest available timings. The voltage was still at 2.6. Keep in mind the highest MHZ rating this board has ever yielded was 191 MHZ with Corsair XMS 300 memory, which is designed to run at 370 MHZ DDR.
193 MHZ CAS 2 1T COMMAND-Sandra 150 Run Loop Test - PASS
The system was stable at 193 MHZ, and was responsive to all commands. Our next overclocking attempt was successful as well.
205 MHZ CAS 2 1T COMMAND-Sandra 150 Run Loop Test - PASS
We finally reached the limitation with the memory running at 205. The system was stable at 205 MHZ, and was responsive to all commands. After 205 MHZ we hit a wall and could not continue. Limitations within the system were a definite culprit. The memory ran past its rated speed until the system prevented any further overclocking. Unfortunately, we can't see how far past its rating it will go.
Overclocking Conclusion
Unfortunately we could not reach past the 405 MHZ mark. We were, however, able to overclock more than ever before. I have no doubts that the memory could run at much higher than 405 MHZ given the proper circumstance, but the KT333 chipset has too many limitations to allow for that high of an overclock. Once we get a KT400 mainboard we will be amending the article.
Now that overclocking tests have concluded, let's just see how fast this ram is...
|