New Posts  All Forums:

Posts by Usario

It seems you're right, AMD has quoted no figure for the PD chips. Though they did for Bulldozer, and they ranged from 61C for 125W parts to 71C for 95W parts. I'm not adamant about 62C exactly, but getting into the upper 60s and above is definitely going to accelerate degradation.We can't be sure he needs more VRM cooling until he gets back to us with regards to the VRM temperature.
Never seen such a ting myself but I'll take your wordAMD's maximum recommended temperature of 62C is with the core temperature measurements... regardless of whether it's actually 62C when it reports 62C, that's what AMD has taken into account when calculating that figure. I've heard actual AMD employees say not to look at the socket temperature, just make sure the core temperature is below 62.What is it lowering the multi to? If it's going down to like 7x you're probably...
All that data is useful, yes, except the socket temperature. As far as OP is concerned I think socket temperature should be ignored completely unless there's throttling. It's unreliable and unimportant. Core temps are what matters. And, when getting into higher voltages, VRM.As far as throttling at 62C on the socket, I'm not sure about the R2.0, but on my original M5A97 EVO I tortured an X4 955 at over 75C core and had no issues (besides degradation of the chip of course)
In your experience, I mean.If you want to, there's no reason not to. Monitor your temperatures and make sure you're not going much over 62C. Voltage only actually destroys chips at levels that you can't reach with air or regular water.If you find yourself reaching unacceptable temperatures and still can't get it stable at 4.4GHz, bump the multiplier down a notch and increase the bclk to 205MHz and see if it works then. With such a miniscule bclk increase you won't have to...
I don't think it's much at all to read up on but whatever. And you are right that the benefits of bclk overclocking have decreased significantly with PD; some chips/boards don't like even going past 205.With an FX? CPU/NB overclocking used to be essential in the PhII days but I recall it not having anywhere near the same importance with AM3+ due to the improved IMC.Theoretically it helps with latency and bandwidth from anything PCI-E, since PCI-E is connected to the NB...
As long as your temps are in line there should be no problem with raising voltage... remember that AMD sells FX-9000 series chips that often do over 1.5v out of the box... the old rule (which, IIRC, was derived from AMD spec sheets) used to be that as long as you're under 1.55v you're good. The only reason to be uncomfortable with a certain voltage unless you're going subzero is if it causes the temperatures to be too high for your liking.This can actually be pretty...
Piledriver can take 1.58v with no problems whatsoever as long as it's cooled well... I've seen people go up to 1.63v for 24/7 use and over 1.7v for suicide runs on good water with no issues. My chip is over 2 years old and no noticeable degradation (though I have had it running on lower voltages for quite some time -- around 1.48v before I got my water loop set up, and around 1.53v in July because the ambient temps disallowed me from staying at 1.58). Even on a more...
First off: you shouldn't be leaving anything on default, unless it's something that doesn't matter at all like SB voltage or something, as "default" often isn't the actual default value, but rather whatever the motherboard feels like setting it to. Set things manually, even if you leave them at the actual default setting. If you want to keep your bus at 215MHz that's fine; mine is at that same speed and often times you can get a higher overclock by a mixed bus/multi OC....
This means one of two things. 1) TSMC has gotten even worse 2) AMD desires bankruptcy After what happened with their 32nm process I'm hesitant to trust GloFo to deliver on pretty much anything.
blargh
New Posts  All Forums: