There is no way to say his cpu is going to last over 6 years it could fail tomorrow with or with out over clocking. and 75c is incredably hot. its a electronic componant running it 24/7 like over clocking will shorten tis life. Course its not gunna burn out just because its never turned on unless you let it run at stupid temps.
The CPU manufacturing/screening process is complex but not random. There are extensive protocols to detect defects and keep them in control. It takes a real fluke for a CPU to pass screening and then fail within days at stock specifications. In fact I've never heard of such a case. It's always been some user error or clear product mishandling.
While there are variations in the tolerances of CPUs that even come out of the same bin, I wouldn't worry about the fluke case of the CPU that burns up at 70C at 1.4V. It's just too unlikely to happen with the current 65nm process.
wasnt saying it would burn up or that the heat was the flaw. I was only saying you cannot say how long a cpu is going to last or any electronic componant for that mattor. but 70c is still to hot who runs any cpu at that temp anyways. The OP is talking about a over clocked system also not stock.
rwaritsdario can roll his eyes at me all he wants but unless he has some super natural power or just knows the exact life span of electronic componants indevidualy he cannot predict it will last 6 years either.
The only thing that can be said is the % of failer in a given time.
My point was telling someone they can gurantie thier cpu's life to 6 years or over if they tamper with over clocking to someone who IMO doesnt seem to know much about it all is missleading.
Since you use TAT maybe you can answer a question. How different is running TAT at "workload level 100%" than prime95 on stressing the systen with the exception that prime95 tells you if there is an error.
TAT seems to stress the cores more than everest or orthos or prime95, that's why my TAT load temps are always 3-5C higher on average. TAT stresses cores to moniter the temp changes - prime stresses cores, ram to test stability. (Correct me if i'm wrong, but these are mine observations) 8)
There are many other condiserations, which I'm very surprised that no one has yet mentioned. Here are four that come immediately to mind:
(1) Moving parts such as fans and spinning hard drives wear out. I'ts a shame to come home to a dead computer because the CPU fan took a dump during our absence, or a hard drive spindle wobbled or ceased during an idle head movement.
(2) Power hits or fluctuations from thunderstorms and overloaded power grids can kill computer electronics dead. Surge supressors won't completely protect you, and how many of us have our systems on a UPS? How many have both? When is the last time we checked the fault light?
(3) Dust accumulates everywhere and clogs CPU and chipset heat sinks and fans, reducing their efficiency. How many of us pull the covers periodically to perform PM's (preventive maintenance) on our systems, or our less knowlegable friends and family?
(4) Power consumed needlessly is a huge problem in this age of global warming. Here in the USA, we represent 5% of mother earth's population, yet we consume 25% of the energy. We won't even bother to inconvenience ourselves by turning off the light when we leave the room. What example does this set for others? For our kids? Something's drastically wrong with this picture. Things have to change.
4. i turn off my lights yes i leave it on along with the other 5 computers on my network. want to drasticly reduce pallution? tell the government they should have allowed the building of electric and water consuming vehicles decades ago when the tech was out instead of saying no because it would kill oil companys.
Seriously though if it wasnt for the fact the rich always want to be richer we would be either at or nearing the end of this pallution energy dependance. If you want that to change this sure isnt the place to do it.
Actually <85c is what Intel tells you to run by, but they measured it in the IHS, wich is about 60c. Soo <75 is rather conservative just to completly avoid throttling*.
As for the voltages, lets just say ive sticked with the big guns in these forumz who had taught me quite some.