DDR2 at 500 MHz for OCZ 1rst test of the Socket AM2

Février 2006 L M M J V S D 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 You can read previous news using our archives. [ Print all news of this day ]

<< Previous day's news February 21, 2006

Next day of news >> February 23, 2006

DDR2 at 500 MHz for OCZ   Posted on 22/02/2006 at 20:27 by Marc

	OCZ Technology announced the release of PC2-8000 Gold Gamer eXtreme / XTC Edition (whew!) memory modules. Available in 512 MB and 1 GB edition they are clocked at 500 MHz with timings of 5-6-6-15 and function at 2.1v. OCZ isn't the first one to announce such memory modules. Corsair already did last May with timings of 5-4-4-9, but because of the low amount of chips that support this configuration, Corsair had to stop this product almost immediately after its release. Anyway, we have to keep in mind that with a Pentium 4/D there is much use in using these modules. They would be more interesting with the Athlon 64 Socket AM2 that should natively support DDR2-800, but if we take a look however at the timings available nothing is less sure.

1rst test of the Socket AM2   Posted on 22/02/2006 at 11:07 by Marc

	Tom’s Hardware has tested an Athlon 64 X2 4800+ Socket AM2 that mainly singles out from the previous Socket 939 version because of the DDr2 support. This sample only supports DDR2-667, whereas other standard models will have to wait for the DDR2-800. Performances have been measured, according to Tom's, with DDR400 2-2-2-5 and DDR2-667 4-4-4-12 memory. It wasn't possible to change memory timings with the preproduction Socket AM2 motherboard. Applicative performances recorded under those conditions were equivalent, or slightly lower at worst (1 to 3%) for the AM2 platform. WinRAR, which is very sensitive to memory timings, is the one that is the most impacted. Changing to DDR2 reduces performances by 10%. If we look however at the memory bandwidth raw performances, we see that they are identical between the DDR-400 and DDR2-667, even under Sandra and despite the fact that this test isn't much influenced by memory timings. So, this is the reason we thought that memory wasn't probably working in DDR2-667 but rather in DDR2-400 because of the preproduction processor problem. That would explain the performances! We have to hope that it is the case, because bad performances additionally to the CPU cooler fixation system modification that will oblige some users to buy new ones won't help the Socket AM2 popularity at least in the beginning.