Originally posted by: coolpurplefan
Originally posted by: taltamir
so... in benchmarks... how much does your performance increase going to 1T?
I don't know about DDR2, but in DDR1, it was something like 3%...? lol
I never measured it as a percentage of base -- but in MB/sec.
I think that with 2T, I had a score of maybe 8,900 -- this was at 1:1 with tight timings and at or below DDR= 720. With 1T, I think it went up to somewhere between 9,200 and 9,400.
Eventually, when I started experimenting with dividers <> (1:1) and set it to 4:5, a combination of overclock to 875, timings at 4,4,4,10 and 1T put the bandwidth something close to around 9,800.
Here, I'd been running the Crucials at 2.175V, but the reported value was something between 2.22 and 2.24V. And the Ballistix went "south" on Christmas eve, when I had the machine loaded up with a demonstration game for the fam-dam-ily.
Luckily, I had a set of rebadged Tracer DDR2-800's, a database of overclock settings, and my wits. So when the fam-dam-ily convened for dinner that evening, all was happy-joy and wonderful.
But IdontCare has a point. If the subject OCZ's spec'd to run at 1T and the Intel boards (mentioned as they have been over the last year ) don't allow adjustment of the command-rate, then you wonder if nVidia boards will allow "auto" adjustment between 1T and 2T, whether that happens with the Intel boards.