My guess is that either AMD or INTEL cpus are good for a lifetime. I expect the problem is something like a capacitor on the motherboard or the Power Supply Unit going bad or some component fried by a power spike.

I have seen errors in applications while using an AMD processor. It was when I experimented with minor overclocking. Even just 5% to 7% overclocking introduced errors and stability problems using AMD even on a liquid cooled block.

I am against any form of overclocking on any system for this line of work and any other.

The Intel system that failed was run under normal specs except for 100% cpu. Some of these errors are hard to track down when you don't have extra hardware to be swapping out.

My plans are just to eventually rebuild her a dual core based AMD based system now that prices have gone down alot.

According to AMD thier systems are "super scalar" in design. Meaning they use "direct connect" architecture. Intel has a serious bottle necks in their Front Side Bus(FSB). I've use Intel system at school/friends and sometimes they will freeze or hang up for a couple seconds while the FSB is overloaded. I never have that problem with my AMD based system.

AMD processors as based on the "Alpha EV6" bus design. A little known and understood supercomputing chip design. I think AMD will do good with the upcoming quad core design as the bus can be directly scaled.


Thanks, it may very well be a capacitor on the motherboard but I don't know.

rbolo29,
Overclocking is an delicate science. You need to be very familiar with your hardwares' abilities and limitations and know you motherboards BIOS inside-out. It has to be done on the proper hardware as well. All the parts have to work as a system: motherboard, RAM, CPU, power supply, cooling. Any weak links will cause instability.

In the example you gave, my guess would be the errors were coming from the MB or RAM. I dont think there's been a CPU made in the last few years that won't overclock 5%. Intel/AMD underclock their chips by a good margin so they can assure years of trouble free operation for the end user. The same can't be said for budget motherboards and value-RAM though. I've seen cases where even the slightest overclocking would cause them to become unstable.

I have 2 water-cooled AMD dual core Opteron 165 based systems. Stock frequency is 1.8GHz, I have them running at 2.8GHz for an OC of nearly 60%! Each has crunched 2000+ WU's without a single invalid result. Supporting hardware includes: DFI LANparty MB, OCZ Platinum RAM, 550w power supply, Swiftech water cooling. If I had cheaped out and bought a budget MB or value-RAM, these results would not have been possible.

I have no clue as to your experience in the OC field so I mean no offense by the above post. If I'm "preaching to the choir" here just let me know.

Thx for that link OzyLynx........ran CPUZ and compared with the table and am glad to see that, though running considerable hotter than the temps u mention, its still substantially lower than the temps that INTEL gives......4.7 yrs old, I'm not speculating on it joining the dearly departed anytime soon.....still typing while arms are resting on 4cm thick wooden desk

ciao

PS, only extra measure is Speedfan......which has always read-out and logged 2.53ghz. Never knew it could throttle by itself. Temp is 34-35 but only when absolutely idle.

Sometimes that's OK.
Just remember the max temps in that list are MAX.
These temps mean almost instant burn-out, no questions asked! In most cases some damage is done at lower temps. Ya takes ya chances.

Good news is, that a simple case fan to draw the heat energy laden air, out of the case more quickly, is a very cheap and highly efficient remedy.
There's not much point to blowing hot air around in your box.

Cheers. ozylynx