6-Way 1GB DDR500 Round-upFri, Sep 19, 2003 - 9:00 AM
So after spending over $2,255 what have we found out? We found that the more money you spend on memory does not mean that you will receive better performance from the modules you just spent too much on. Below, we have written in-depth conclusions for each company as they all need some individual attention, so you don’t end up wasting your money!
How can anyone ignore the fact that the 1GB OCZ EL 4000 GOLD that we paid $470 on was out performed by the OCZ EL 4000? I simply cannot overlook that. The 1GB OCZ EL 4000 was $110 less expensive and it out performed the GOLD series! OCZ Marketing claims their GOLD line is a better product and tries to back this up by stating that it goes through one additional step of testing on an ABIT IC7 motherboard and it also comes with a different SPD (lower timings). Our retail bought modules showed that this was not the case after benchmarking and stability testing. Could we have gotten a high quality/lucky EL module? Sure, I am not ruling that out, but if they are marketing their GOLD series as being a better product I expect to see a better product when running benchmarks.
Once you look past the marketing and pricing issue the OCZ 4000 series has proved to be the highest overclocking DDR500 modules out of our test group for both loose and tight timings. We therefore can conclude, if you want to run the most aggressive timings and highest speeds possible then the OCZ 4000 series modules are the dual channel kit for you to pick up. Just be careful which module you buy, as naming and how much money you pay has shown not to matter from what we have seen.
**UPDATE**– In our review we found that OCZ PC4000 EL ran faster than the Gold, but there is the possibilty that we got a lucky stick. Therefore, don’t expect that every stick of PC4000 EL to be exaclty like this one. After a dozen or so e-mails from readers and being contacted by OCZ we want you to know that OCZ again told us that additional testing does go on.
For current pricing on various OCZ memory modules check here.
Kingston has had some great success with their HyperX line that was released back in November 2002, but recently made the bad decision to use Samsung IC’s for their HyperX 4000 series. This limits the overclocking potential and is why it is not possible to run lower than CL3 on these modules. It should also be noted that this was one of the modules that had memory errors when run at default speeds and timings when we used memTEST86 and the RSTPro. Due to the poor IC choice, and memory errors we best advise our audience to avoid the Kingston HyperX 4000 series if in search of solid performance beyond DDR500. I am sure that Kingston R&D is upset with their choice and will prevent poor IC choices on their next series of HyperX.
For current pricing on various Kingston memory modules check here.
We have never used Geil memory before, but due to the popularity of the brand in various forums and their low price point we were interested in them. Based on this, we picked up both of their PC-4000 series to see if the hype was true or just more “insider” marketing! The more expensive and nice looking Geil Golden Dragon Series memory proved to be a memory error trap. We found the memory to have stability issues when running many applications and benchmarking programs at default! Memtest86 and RSTPro testing confirmed our desktop issues and showed many errors at default and overclocked settings. It may look pretty, but the Geil Golden Dragon Series was the worst performer in our group as far as stability is concerned.
The other module we had from Giel was their Platinum 4000 series. This module proved to be a solid memory module at default and even generated no errors on memtest86 and the RSTPro testing. When it comes to overclocking we had some success at loose timings, but the module did not like running much tighter than the default CL2.5 timings. From the testing we have done Geil Platinum would be fine for a user that would like to run DDR500, but not into overclocking or extreme performance beyond 500MHz.
For current pricing on various Geil memory modules check here.
Out of the six modules tested we found that the Corsair XMS 4000 memory line was the most stable of the bunch. It had no errors from memTEST86 and RSTpro testing when ran at default/aggressive timings at DDR500 and even at high speeds at loose timings! No other memory series we tested can boast these results. As far as overclocking is concerned, the Corsair XMS 4000 has ample head room, which allowed us to reach speeds above 520MHz on our module even when run at tighter CL2.5 timings. The fact that Corsair took a conservative approach on setting the timings even though the memory runs fine at CL2.5 actually plays in favor of the end user. Corsair makes it simple as there is only one DDR500 level of performance to pick from (XMS and PRO series are both 3-8-4-4). The Corsair modules are able to get an extra boost of performance that is available thanks to the Hynix IC’s and the great R&D work that Corsair put into the actual PCB design. With a solid price point, no errors on three of our four testing stages, and good performance it is our belief that the Corsair XMS 4000 series is the #1 memory pick on the short list of DDR500 memory modules.
From the conclusions we found on this review and working with each of these modules for over a month we feel that the 1GB Corsair XMS 4000 dual channel kit is worthy of having our editor’s choice award as our pick for the best of the DDR500 memory currently on the American market or possibly even the world.