New Posts  All Forums:

Posts by Sam OCX

I thought for a long time that all OCZ 1600 7-7-7 1.65V rated kits are BBSE but recently I've seen a set with a completely different IC shape so the statement no longer holds.OCZ were very chaotic with their IC selection, you can never be fully sure about what you're going to get.Gigabyte Z87 boards are not good at clocking RAM.I have a set of Samsung that does 32M at 1466 9-12-12 tight on ASUS and even MSI, but won't even POST at 1466 9-12-12 or 9-13-13 on Z87X-OC (1400...
Two of my sticks can do this: but I guess you're running secondaries tighter than (equivalent of) 4-88-10-4-24-4-5-8 which I used for the run. I never bothered to test tWCL 6 on Ivy since I thought that it would need too much voltage and compatibility wasn't perfect. Unless the kit used for 1200 7-8-8 run is out of this world then the difference between tWCL 8 and tWCL 6 shouldn't be too big.
The only difference between RipjawsX, RipjawsZ, Sniper, Tridents and TridentX models which have exactly the same specs are the heatspreaders and associated marketing. Same goes for G.Skill 8GB modules rated 1600 7-8-8, 1866 8-9-9, 2133 9-11-11 and 2400 10-12-12. All of those are based on the same Samsung 4Gbit B-rev chips, will easily do each others' specs and overclock roughly the same.
Can you test if Hypers need as much voltage for CL7 + tWCL6 as they need for CL7 + tWCL8 at same frequency? (not the case with PSC and BBSE, obviously)
George, how much volts for 1200 7-8-8 on air?
Left vs right tertiraries will gain you about 0.6s on average. The Efficiency Score MemTweakIt pops out does not actually represent how fast memory settings are since it depends on not-so-important Refresh Interval (which you actually raise under benching) and doesn't depend on some of the secondaries. Pi times done while memory binning are not directly comparable as I sometimes run/attempt the test multiple times in a row with different process histories.
1.58V won't kill or damage the memory in any way, so there is nothing for you to worry about.
I can't see anything seriously wrong with A-Data XPG V2 having reviewed a couple of their kits. Performance-wise, A-Data and Kingston will do the same. If Kingston costs $20+ more for the same specs, I would go for the A-Data.
There is about 0.2s difference in SuperPi 32M between default values (8-11k) and manually set 25000. The latter might also require a tiny bit more voltage.
Retesting some PSC and BBSE numbers with 4.5GHz core/cache clocks:
New Posts  All Forums: