New Posts  All Forums:

Posts by MaxT

There is agood chance you will kill your CPU. I've done it on my 3770k successfully, but only since i had experience doing this with older opterons (which i killed two of ). It's not so much about the die itself, although you have to be careful not to chip it, it's about not cutting the PCB. If you do, you can end up with anything from a gimped memory controller (one dead channel) to a dead chip.
from what i've read, you have to read your WHEA error messages to see if they are pointing to the CPU or to the memory. If they say a CPU core fixed an error then add more vcore. If it points to a certain DIMM that means you need to stabilize your memory. manual vs offset voltage probably won't make much of a difference in idle temps. it more depends on whether you still have EIST and C1E enabled. If they are enabled, the CPU will run at much lower clocks/multipliers,...
I want to stay with a good mix of MHz vs tighter timigns. 2200 9-10-10-24-1T is what I would like to get full stable 24/7. Every guide I read recommended setting VCCIO and VCCSA. Took me 2 weeks of testing to find out they do nothing for my overclock . Now I am just trying to get the tight timings stable at anything under 1.6v. The chips on these sticks can take much more, it's the PCB i am worried about. Either way 1.6 is my limit, at which point I will start relaxing the...
Running mine right now at 2x4GB @ 2200 9-10-10-24-1T, with the secondary and tetriary timings tightened up at 1.56V. Still testing with Prime95. Might have to go a bit higher on the voltage. Chaning VCCIO and VCCSA did nothing for stability for me. Only cpu voltage and ram voltage makes a difference right now.
124 is not always vcore, sometimes it's VCCIO (VTT) or the memory voltage Are you overclocking your memory at all? Were there any WHEA errors in the event viewer?
I found it doing more harm than good. The premium I paid was for extra cash and what seems to be a better binned chip. I can always reenable it, but for now I see no reason to. Games don't benefit, and it seemed to hurt the GFLOPs numbers in the synthetic benchmarks, so who knows what it does in real world.If i find a reliable way to benchmark BF3, I will test it with HT on and off to see if there is any benefit to it.
Here is my submission!
Maybe. No way to know for sure yet, Ivy has only been out for a few months.Multiple sources said not to go over 1.35 for 24/7. That may be concervative, but I wanted to stay under 1.3 for peace of mind.
still WIP this is the current GFLOPS in Linx AVX with HT disabled (gaming PC, no need for it right now) and the 4x4 ram at 1866 9-9-9-24-1T. Still testing for stability and will be pusing the memory higher. The goal is 2133 9-10-10-24-1T. By the way, disabling HT did nothing for the temps for me. It runs maybe 1 degree lower. But at least 14-15 GFLOPS gained in Linx.
If you get like 1 a year then no. If you consistently keep seeing them in the log, then yes. They shouldn't be there.
New Posts  All Forums: