What voltage are you putting through your chipset!?

[Elaruwan]

And your LDT?

 

I don't see any point raising LDT, and theres no point going over 1ghz, and since memory controller is off the chipset, no point in raising chipset voltages.

 

Some people say they get a more stable clock with a slightly raised chipset voltage.

 

Just wondering because my 170 is flopping all over the place. 1.425v primed for 8 hours at 2.7ghz, was running a series of primes between 270 and 275, to see what else i could get out of it at the same voltage.

 

When i found i couldn't get 272 to run stable for more than 50 minutes, i decided to go back to 270 (just for thrills) and prime failed after 51 minutes. Crazy.

 

So primed now at 1.45v and its fine (8hours).

[sparkison]

Chipset Voltage Control

This is the voltage applied to your motherboard chipset. The stock voltage for my chipset is 1.5v, but currently I’m running mine at 1.7v just to make sure its getting enough juice since it runs pretty cool for me anyway. I haven’t noticed much of an improvement in stability with upping this voltage as I did when I had a nforce2 motherboard but I like to up it a little anyways.

 

This is from the ultimate overclocking guide here. So to answer your question, I dunno why you would raise it? Sorry

I honestly have not noticed a stability difference from stock voltage to higher voltages.

And it doesn't seem to affect the score of benchmarks either. So my suggestion would be to leave it stock so that it runs cooler, unless you are having stability issues, then try raising to see if it helps.

[Elaruwan]

Yeah, I've read that, and back the days of NF2 it was cool to raise voltage on your chipset, I had to do it on my Lanparty ultra rev.b to get higher FSB's.

 

Just pulling things out of thin air really, I can see why theres an option for me to raise my LDT voltage, (just incase i want to O/C my HT for some reason), there must be a reason for me to up my chipset voltage!

 

And no, adding more heat to my room does not count as a valid reason

[CPDMF]

It didn't help me any so I just leave them at 1.5 and 1.2. If it did help I wouldn't have a problem turning them up though.

 

It's pretty common for me that prime will pass at a certain clock one time and then fail, it just means you were right on that edge.

[joncrap]

i've seen no difference in stability when uppin my chipset voltage

[madu]

Upping the LDT voltage helped me. I am slihghlty over 1G HTT and I was not prime stable with my current clock. then I upped the LDT voltage a notch (1.3v) and it passed prime.. I also upped the chipset voltage a notch too.. maybe that helped too..

[UncleDavid218]

Stock.

 

There's no point in raising LDT voltage other than raising temperatures.

[CamB]

Buried in some post about memory timings I found a reference to raising LDT voltage if CPU-Z gives an error when loading. I didn't see that as enough of a reason!

 

I also found that at really high HTT (like 300) I needed to raise my chipset voltage (otherwise I "lost" the onboard audio and one of the NICs), but even then it only helped a bit. My CPU (or my skills) is always the problem.

[UTLanpartied]

defaults, nothing extra

[red930]

Upping the chipset voltage to 1.9 volts maxed out my FSB at 389 or so, I leave it at 1.8 since it doesn't really affect heat for me.

 

I also leave the LDT at 1.5 just 'cuz I see no reason not too, never noticed a heat difference or anything, so I don't worry about it.

[twivel]

I was able to break 340 FSB without increasing either Chipset or LDT volts.

 

Ok, So I didn't used to believe in burn-in, but I'm not so sure anymore. I think chips get used to running at a certain voltage - and large voltage change can be less productive than gradual ones.

 

When I started testing my overclock of the CPU with new memory, I did a slow, progressive climb as suggested by others on this list. I spent some time mpriming at each step. When I got as far as I wanted to go with voltage(1.6+-), I decided to back down to an earlier speed/voltage combination that I found before. But when I first backed down, it wouldn't run at that speed/volts combination. Only after mpriming for a while at lower speeds/voltages was I able to achieve the same results again.

[cuoficr]

If i raise my LDT at all I max out allowable error's in prime95 after about 30 seconds of the test running, although im not a huge fan of P95 that scares me so i leave it stock, and i think my chipsets at 1.7, dunno why just bumped it out of boredom i guess.