New Posts  All Forums:

Posts by Sam OCX

Final proper tight runs on BBSE before my last set leaves for new owner:
Testing another set of Kingston DDR3 MFR: 1000 / 7-10-10-24 / 1.88V 1100 / 8-11-11-24 / 1.79V 1200 / 9-12-11-27 / 1.72V 1300 / 10-12-12-30 / 1.68V 1400 / 11-13-13-35 / 1.64V 1500 / 12-14-14-35 / 1.65V 1600 / 12-15-15-36 / 1.80V
The kit is not fit for 32M at 1400 on air, one of the modules just can't do it. The other is fine, however - it does 32M at 1466 8-14-8 easily.This is the actual kit that Xtreme Addict ran 32M at 1400 6-9-6-21 before/at Computex. Since then, some BGAs might have got loose so this set was giving us all sorts of problems (from no POST on air to hard coldbug at -5c). Couple of weeks ago XA did some soldering work on it (so set looks like a bloody mess with all sorts of pastes...
Testing what's left of my old faithful 2000C6 PI.
Yes.Put simply, Team 2666C11 will easily do G.Skill's 2400C10 spec while the inverse is most likely not true.
There shouldn't be any difference between Red and Black versions apart from the colour. Spec is the same, lineup in the same. Won't be the first time when G.Skill offers same mems in different colour (like Arse, for example).
Using same version of XP 32 that I've been using for years. My efficiency is far from perfect, I didn't use a dedicated OS for wPrime, just "tuned" a two month old stability test OS.Be sure you don't open any software (like cpuz) before or while you run the benchmark. Also check your C-states (C6 especially). Sometimes you get phantom CPU load when not running anything, this also dramatically worsens the benchmark scores.
Is tRAS 15 a must at 1500c12 with Hynix, or you're just following everybody else?
Roy, how is the performance for 4x4GB vs 4x8GB at same clocks/timings?
week1435 2400C9 Tridents:
New Posts  All Forums: