Conclusion
The PDP Patriot Extreme Performance PC2-6400 Compare Prices on PC2-6400 Memory holds a commanding position based on its overclockability alone, and an ultra-low final price (around $200 after $50 mail-in rebate) is astounding. The rebate ends February 28th, with no word yet on future rebates.
Next up is Buffalo FireStix. Available for around $240, and just a tick behind the Patriot Extreme Performance in overclocking, its price comes without with the hassle and uncertainty of mail-in rebates. These may actually overtake Patriot Extreme Performance in value once PDP Patriot’s rebate expires.
Mushkin SP6400 and Wintec AMPO both exceeded our expectations for standard performance modules. Both are designed for ultimate compatibility and stability rather than speed, but both are still competitive overclockers. Mushkin’s slightly higher achievable speeds are completely due to a slightly better tolerance for insane voltage levels, where Wintec AMPO leads at the more reasonable 2.10 V setting. Fast enough for all but the most aggressive overclockers, a lower price could keep the ball in Wintec’s court unless you actually need factory support : Mushkin offers toll-free tech support while Wintec lets you pay your own long-distance telephone rates.
Only four out of the ten kits tested met all of the criteria for final bang-for-the-buck consideration, and of these, the rebated PDP Patriot Extreme Performance is both the fastest and cheapest. Wintec AMPO offers the highest value for buyers who don’t use or qualify for rebates, while Buffalo FireStix and Mushkin Standard Value fall in the middle for both speed and price.
What About The Rest ?
The most common criticism in large comparisons is the lack of a few popular brands, but rest assured that I didn’t drop the ball on this score. Four brands missed the review submission deadline due to such common occurrences as product line updates, marketing department changes or miscommunication.
Another concern may be the lack of game/application benchmark results for those programs used in stability tests. Why not grab the numbers ? Each change in RAM speed required a similar change in CPU speed, a change that would have unfairly slanted the results, given that most overclockers will never reach the maximum frequencies of these modules at a 1:1 FSB :DRAM ratio. If you’d like to see the benefits of CPU overclocking using a fixed memory ratio, please refer to our CPU Overclocking articles.
Finally, there’s the question of timings versus frequencies. At 8-12 hours of testing per module set, performing a second set of tests on ten modules using the same platform would have held the article back by over a week. Any long delays in this fast changing industry could have caused this article to lose its relevance, so I had to focus only on overclocking ability for now. A Core 2 platform update to our previous article "Tight Timings vs. High Clock Frequencies" may be forthcoming, depending on reader response. Onward to the Forumz !
