Thanks Max. I didn't know it's called silicon, without an e :Straight Face: I was just trying to give a simple answer though, as he was going for a little more than just a turbo...

Thanks Max. I didn't know it's called silicon, without an e :Straight Face: I was just trying to give a simple answer though, as he was going for a little more than just a turbo...

Right, my point though is that these processors are in fact designed to have their clockspeed adjusted upward and as long as you keep power draw within the design specification you won't be shortening the life of the CPU at all. For some (modern) chips this upper bound may be 4GHz or more. Also, with the use of exotic cooling which reduces the component operating temperature below ambient temperature (AKA sub-ambient cooling) one can push this upper bound even further without shortening the life of the processor. To a lesser extent this is also true of near-ambient cooling solutions (like H2O).

Thanks Max. I didn't know it's called silicon, without an e :Straight Face: I was just trying to give a simple answer though, as he was going for a little more than just a turbo...

Not that there's anything wrong with having some "silicone" in your life :-D

I've only heard of a few cases of a CPU being the first component to die. If you've overclocked properly then the CPU will probably outlast the hard drive and motherboard

Thanks Max. I didn't know it's called silicon, without an e :Straight Face: I was just trying to give a simple answer though, as he was going for a little more than just a turbo...

They are easy to tell apart: Silicon is hard, silicone is soft! :Big Grin:Cheers,- jahman.

Any overclock is detrimental to the life of the chip. Any overclock is decreasing the life of the chip regardless of voltage or temperature. A transistor is a switch designed to open and close at a certain frequency; exceed that frequency and you are causing wear. Transistors are also sensitive to voltage regardless of heat. Other components on the board are susceptible to heat and are not monitored (you decrease the life of more than only the CPU). Intel guarantees the chip for 3-years (damage by overclocking is typically undetectable and having warranty denied due to it is slim to none). Intel put much work into analyzing the life of its processors so it is highly probable a non-clocked chip will last more than 3-years. Overclock it and that probability diminishes; to what extent nobody knows; maybe many years; maybe hours or even minutes.I am prepared to accept the risk of my 4.5GHz clock based on the specifications Jahman referenced.

Any overclock is detrimental to the life of the chip. Any overclock is decreasing the life of the chip regardless of voltage or temperature. A transistor is a switch designed to open and close at a certain frequency; exceed that frequency and you are causing wear. Transistors are also sensitive to voltage regardless of heat. Other components on the board are susceptible to heat and are not monitored (you decrease the life of more than only the CPU). Intel guarantees the chip for 3-years (damage by overclocking is typically undetectable and having warranty denied due to it is slim to none). Intel put much work into analyzing the life of its processors so it is highly probable a non-clocked chip will last more than 3-years. Overclock it and that probability diminishes; to what extent nobody knows; maybe many years; maybe hours or even minutes.

No, it isn't. Turbo exists because chips do not usually run at their maximum rated capability, turbo kicks on and pushes the chip up to this level. The newest 32nm Sandy Bridge Intel chips even go beyond this level for brief periods of time. Intel guarantees their processors for a period of 3 years while running within normal operating conditions. This includes running @ maximum TDP in a 24x7 operating environment.

Life is short, OC your chip!EDIT: Max, that's twice you have ninja'd me

No, it isn't. Turbo exists because chips do not usually run at their maximum rated capability, turbo kicks on and pushes the chip up to this level. The newest 32nm Sandy Bridge Intel chips even go beyond this level for brief periods of time. Intel guarantees their processors for a period of 3 years while running within normal operating conditions. This includes running @ maximum TDP in a 24x7 operating environment.

Firstly, are you meaning absolute maximum voltage, or upper end of the functional operational limit when you mention 'max rated capability'?I'm thinking if you take a million chips and subject them all to varying levels of voltage from the lowest in spec voltage to their max 'in spec' voltage, the ones at the highest end will fail earlier, and it will be a statistically significant difference. Put another way, if I don't subject it to ANY voltage, it will last indefinitely from a human reference point, gamma ray bursts aside! If you overvolt them by let's say 20%, you can expect much more failures, and the rest will scale fairly predictably in between. All that aside, individuals mileage will vary ;o)In the real world where I visit now and again, this won't make a difference anyway. Chances are excellent that you will be able to OC your chip, even outside of the functional operational limit to some degree, and get away with it, most of the time. As the other poster stated, these things are engineered to do better than barely adequate. They need to have this tolerance built in else they'd fail more often, because I'm sure other components involved in voltage regulation may not always work perfectly either!Noel

I have one golden rule about overclocking, here it is....if you can't afford to overclock don't!I fried a QX6850 one year after I got it by pushing it well over the rated voltage...($1,000.00).I ran a Q9650 at 1.52v @4.32GHz for two years before I upgraded to what I have now, I've been running this i7 980x @ 4.67GHz with 1.537v on the core 24/7 since 10/21/2010.No offence to all SB owner but the SB overclocking is idiot proof compare to the old way where you had to fine tune your rig like I did with the one I have right now.So if you are up to it and you can afford it go for it but if you can't sleep at night wondering about the life span of your CPU....you better play it safe.Good luck either way.

I have one golden rule about overclocking, here it is....if you can't afford to overclock don't!Good luck either way.

Here, here! That's it in a nutshell, though it clearly implies you're in the camp of acknowledging the implied risk. I myself secretly want to kill my Q9650 so I can upgrade w/o the guilt! Actually that's not true I'm trying my level best to keep this thing running til Ivy Bridge or beyond. My main agenda will be in getting a bigger screen and not having to have that choke the system. I'm still pretty impressed w/ my current machine which is getting quite old now--3.5y and counting. I too killed a QX processor which makes me wonder if they bin them to die young just because they can and get the last laugh on the 'enthusiast'! I'm def more conservative on the replacement I'm using now. That's a pretty fast FSB to get 4.3Ghz or whatever it was. I don't know if this board could hack that. I ran it at 4Ghz w/ softening timings and slower DRAM freq, at 1.3625v, but she caused Prime95 errors quite quickly. Didn't play w/ it much as I don't think it's worth the extra voltage at this point.

By overclocking? In light of some recent posts and the latest sandy bridge/fsx craze, I've been wondering what speed is safe and won't be detrimental to the cpu? Am I ok running my processor at 4.5 or should I tone it down a little? Hoping some experts can weigh in. Thanks.

Nah, it's just plug and play. The faster the GHZ the better it is for trouble free flying on cloud 9! Believe me, would I lie to you? :Big Grin:Just kidding! Sandy Bridge has made some FSXers happy some of the time, but I guarantee you, it has not made any of the FSXers happy all the time! Kind regards,PS: Is it worth it? Yes, of course!

Firstly, are you meaning absolute maximum voltage, or upper end of the functional operational limit when you mention 'max rated capability'?I'm thinking if you take a million chips and subject them all to varying levels of voltage from the lowest in spec voltage to their max 'in spec' voltage, the ones at the highest end will fail earlier, and it will be a statistically significant difference. Put another way, if I don't subject it to ANY voltage, it will last indefinitely from a human reference point, gamma ray bursts aside! If you overvolt them by let's say 20%, you can expect much more failures, and the rest will scale fairly predictably in between. All that aside, individuals mileage will vary ;o)In the real world where I visit now and again, this won't make a difference anyway. Chances are excellent that you will be able to OC your chip, even outside of the functional operational limit to some degree, and get away with it, most of the time. As the other poster stated, these things are engineered to do better than barely adequate. They need to have this tolerance built in else they'd fail more often, because I'm sure other components involved in voltage regulation may not always work perfectly either!Noel

I didn't mention voltage, certainly not over-volting. TDP is power consumption.

I didn't mention voltage, certainly not over-volting. TDP is power consumption.

Just curious, how would you increase clock without increasing consumption? You would have to be a very lucky man to do it without increasing voltage. regardless of voltage increasing clock would increase consumption. To be more accurate TDP is related to heat disipation.

Just curious, how would you increase clock without increasing consumption? You would have to be a very lucky man to do it without increasing voltage. regardless of voltage increasing clock would increase consumption. To be more accurate TDP is related to heat disipation.

Watts and heat are one and the same. Now the choice is yours, you can get your max TDP with GHz or get it with volts. Intel sez max is 95 watts and cpu case must able to maintain 72c or cooler. So choose your poison.