Discussion in 'News' started by D_o_S, Apr 20, 2007.
yeah, that's why im wondering if they can detect it.
I dont believe it, think it is a scare tactic, no. I am french, no.
thanks for the tips on how to pull the cpu heatspreader off, that thing sucks, I hope one day the heatspreader will rip off for me. I was going to try a blow torch but if AS5 does the trick then I will have to give it a nice strong yank. F-that stupid concaved IHS
gates doesnt own intel....
In this case you can only pull it off.
Make sure you pull it strait up so you don't bend the pins.
One time I actually bend a dozen pins on my s939 and need to bend them back manually.
I call fake.
Why would intel implement a feature that would not only cost in developing, but take up precious wafer space,
when the amount of people who overclock is absolutely tiny compared to entire sales.
It does not make sense at all from a business standpoint, as they would spend more money in the development, and wafer space, than they would recoup by refusing certain RMA's.
not to mention the article gives no sources what so ever. So the entire thing could easily have been fabricated.
Like Zubasa said, boxed cooler makes thing alot more complicated, not having the room to twist&shout But to help the process of pulling it without prosessor following do some CPU burning test before you take it off. CPU Burn-in is a nice little proggie for that http://users.bigpond.net.au/cpuburn/ or prime95/any other CPU torture test. 15min should get the thermal paste nice and hot and should help the heatsink come off much easier. Now take side panel off before you start and case already on sideways helps to get the heatsink out as soon as you've stopped the test and turned puter off + taken power cord out.
I too doubt theres a "detection" method for this, aside from maybe a voltage fuse for extreme conditions. (As a surge could possibly cause more harm than a dead cpu)
Along with the fact, Overclocking is based around finding the weakest link. With multiple millions of transistors, you would have to have a detection point for every possible point of failure, making this a nusance rather than a feature.
And would actually, be likely to provide a HIGHER return rate.
No need to. Undervolting isn't a concern. all its going to do is be unstable, and continue to work at proper speeds/volts. If you kill your CPU from "undervolting" you're only fooling yourself.
Who cares? Nobody kills CPUs unless they are idiots, and idiots don't deserve new CPUs!
At best, i think people were right on the fuse count.
say stock volts was 1.3V - if the CPU hit 1.5V or higher, the fuse goes, without stopping the CPU working
If a warranty came in that looked suspsicious, they could go to the effort of checking it out. but with something so small, its not financially viable to pop out an electron microscope or whatever for a single CPU. (or however they test it)
Separate names with a comma.