Tony, <I thought Intel wasn't changing the 820 chipset for purposes of 3 RIMMs.>
I don't understand the nature of the three RIMM problem, but I can see how a new version of the 820 chipset can help workaround physical problems on the motherboard.
By the way, the three-RIMM instability issue seems to be one of those really hard-to-find problems. AOpen already has motherboards out in both 2-RIMM and 3-RIMM configurations, and they said they failed to notice any instability issues with the 3-RIMM design. AnandTech also failed to notice any stability issues. See:
anandtech.com
As you can guess, this does limit your memory expansion quite a bit, but almost all manufacturers have moved to 2 RIMM i820 designs. The exception being a member of this roundup, AOpen, who offers both a 2 RIMM and 3 RIMM motherboard solution. According to AOpen, they failed to notice any instability issues with their 3 RIMM design and, in our tests, we failed to notice a difference in stability even when using 3 Rambus modules in the AX6C outfitted with 3 RIMM slots.
Wow, and to think that the whole 820 fiasco was caused by a problem that not everyone is able to reproduce. I'd be willing to bet that if this were a product from VIA or AMD, the problem would have been deemed a non-issue and the product wouldn't have been delayed. Seems like Intel went head-over-heels to ensure a very high standard of quality.
Tenchusatsu |
