jmke

A HEAP of heat problems with Intel’s E8400 microprocessor are being reported on a host of hardware forums. Users are claiming that the sensor diodes aren’t up to scratch and that’s causing the overclockers to get overheated themselves.

http://www.theinquirer.net/gb/inquir...reported-intel

__________________

jmke

that and the fact that overclocked CPUs require more voltage over time to run at the same speed, leading to believe there might be a risk for sudden death syndrome (much like with overvolted Pentium 4 northwoods). 45nm + lots of voltage = short lifespan, diminished overclock over time.

__________________

Kougar

So far it would just appear to be another Tjunction bug, socket based diodes for the most part look to be working fine since they use a different source.

Not seeing any cases regarding overclockers having to raise voltages to maintain their overclock. I think the chips are still way to new and rare to draw any conclusions yet.

jmke

check XS forums

__________________

Sidney

There is no free lunch; if Intel or anyone could produce a CPU at default voltage or slightly above to have 50-100% overclocked speed constantly and sells them at bargain basement price and forever lasting, you are a happy person all your life.

I wonder when Boeing will produce a plane with a slight overclocked engine will fly around the world twice at twice the speed without refueling.

__________________
lazyman

Opteron 165 (2) @2.85 1.42 vcore AMD Stock HSF + Chill Vent II

Kougar

I went back for a 2nd look and browsed some of the Wolfdale threads... to much hype from a couple cherry picked processors meeting cold hard reality that most chips are NOT going to reach 4.5GHz, let alone 4GHz on air. Let alone reasonably stable. I fell for it with G0 Q6600's and once was enough for me, although this time the hype seems worse if anything... And a point needs to be made that the Q9650 is more than a few bin grades of quality better than an E8400. If anything it was highly likely Anandtech and other tech sites got cherry picked Q9650's.

I also noticed very many are putting in 1.5-1.65v into their 45nm processors to reach 4.0 to 4.5Ghz. I would expect nothing less than chip degradation or sudden death.

Heck, all of them are saying "1.45v" is the safe maximum official Intel voltage... which is completely wrong. I downloaded the spec sheet for the E8000 series and whomever bothered to read the paragraph above the "min/max" votlage table would know that Intel specifically warns of chip degredation when the normal VID range is exceeded. Anything over 1.45v is just guaranteed to lead to damage, that is all that number means. It had nothing to do with the VID. Wolfdale's actual "max" safe rated VID is only 1.3625v

Rutar

1.6 Malay P4 Northwood?

Sidney

I am still running 2.0 nothwood at 3,2Ghz since 2002 @1.6V with 1.5 default; E2140 1.6G at 3.1G and 1.38v ... but I don't expect it is a daily bargain or free lunch. Air cooling stock cooler/fan!!!

__________________
lazyman

Opteron 165 (2) @2.85 1.42 vcore AMD Stock HSF + Chill Vent II

Kougar

Yeah, still running my 2.8Ghz Northwood somewhere around 3.2-3.4Ghz, I don't even remember. Had it running 24/7 since 2003, was a 800FSB chip though so did not have enough FSB headroom on my Abit IS7, got up to 950FSB and it just couldn't handle more. Stock volts of 1.525. It used to be a Gallatin though, CPU-Z's cache latency program found the deactivated L3 cache long before CPU-Z began reporting it as a Gallatin. Eventually realized I had a 30-caps chip after more than a year after I had had the thing...

Definitely not a "free lunch" compared to power consumption/performance with say an E4500 though. I am thinking of selling it now, even a Celeron E1200 system would be cheaper on power and at least 5x the performance before a nice overclock.