Will Overclocking GPU reduce lifespan?

[AYDANN6ix9ine]

So I have a Superclocked EVGA GeForce GTX 970 ACX 2.0 Graphics Card paired with a Ryzen 5 2600X OC'ed to 4.1Ghz and DDR4 3000 16GB RAM OC'ed to 3133mhz and want to overclock it so there is better performance in gaming but video rendering too and less "bottleneck". If I overclock it with let's say a 150-200mhz+ core clock and 350-400+ memory clock while having to increase voltage and power limit and temperature limit, will it reduce life significantly?

[Jurrunio]

significantly, no.

[Fasauceome]

Voltage is what kills. What's your Vcore?

[Kalm_Traveler]

13 minutes ago, fasauceome said:

Voltage is what kills. What's your Vcore?

and heat

[andrewmp6]

If you keep it cool it won't hurt it at all.

[Mick Naughty]

Heat kills, as does poor quality components. 

 

An overclock itself won’t do anything. 

[xg32]

3 hours ago, AYDANN6ix9ine said:

.

iirc 970s are locked to very low voltages around 1.3v, so it won't be a problem, just watch temps below 80-83 prefered, but technically oc'd parts should die faster than non-oc'd parts on paper.

[Oliver Murphy]

If you proper cooling and do overcooling properly then no it won't impact on your Graphic card lifespan

[Enderman]

7 hours ago, fasauceome said:

Voltage is what kills. What's your Vcore?

No, that the completely opposite of the truth.

Temperature is what causes damage, not voltage.

 

[Stormseeker9]

7 hours ago, fasauceome said:

Voltage is what kills. What's your Vcore?

Voltage doesnt kill anything. Its the current (amps) what kills a device, component or a human being.

[Fasauceome]

4 hours ago, Enderman said:

No, that the completely opposite of the truth.

Temperature is what causes damage, not voltage.

 

Ah ok I'll just set my Vcore to 5 volts then, I'm sure it's fine.

 

Really though, semiconductors wear out due to the kinetic energy of electrons degrading the internal transistors, even if you kept a CPU at 30 degrees it wouldn't last forever. A copious amount of voltage at a safe temperature will cause problems and a safe voltage at a copious temperature will cause problems. It's not the opposite of the truth, it's only half of what you need to look out for.

 

And yes, I know what popcorning is.

[Fasauceome]

4 hours ago, Stormseeker9 said:

Voltage doesnt kill anything. Its the current (amps) what kills a device, component or a human being.

Alright, if you wanna get super technical about it it's actually Coulombs, which are then composed of electrons, because one Amp is one Coulomb per second. Also, with 0 voltage, any amount of current has no potential and therefore can't move, so you need voltage and current to kill people

 

Still though, there's no setting in the BIOS for "current control" so when someone tweaks voltage to change wattage then that's how they overload a CPU.

[Stormseeker9]

19 minutes ago, fasauceome said:

Alright, if you wanna get super technical about it it's actually Coulombs, which are then composed of electrons, because one Amp is one Coulomb per second. Also, with 0 voltage, any amount of current has no potential and therefore can't move, so you need voltage and current to kill people

 

Still though, there's no setting in the BIOS for "current control" so when someone tweaks voltage to change wattage then that's how they overload a CPU.

Yes the one without the other is nothing.  Correct.

should of left my explain part in my comment, would of saved you from typing it out.

 

 

[Fasauceome]

1 minute ago, Stormseeker9 said:

Yes the one without the other is nothing.  Correct.

should of left my explain part in my comment, would of saved you from typing it out.

Might as well take some fun in nitpicking if I can.

[Kalm_Traveler]

5 hours ago, Enderman said:

No, that the completely opposite of the truth.

Temperature is what causes damage, not voltage.

 

 

4 hours ago, Stormseeker9 said:

Voltage doesnt kill anything. Its the current (amps) what kills a device, component or a human being.

in the context of this conversation, voltage will indeed kill things. No need to purposely mislead someone who is trying to learn

[Fasauceome]

59 minutes ago, Kalm_Traveler1 said:

 

in the context of this conversation, voltage will indeed kill things. No need to purposely mislead someone who is trying to learn

And not even in the context of this conversation, voltage is a form of potential, which leads to kinetic energy, and that's what degrades things over time.

[Kalm_Traveler]

1 hour ago, fasauceome said:

And not even in the context of this conversation, voltage is a form of potential, which leads to kinetic energy, and that's what degrades things over time.

yep - just trying to keep things simple for the OP too, since the end result of our answer should be something to the effect of "running your PC parts with too much voltage is an unwise idea", technical specifics aside.

[Mira Yurizaki]

The only thing overclocking would do to reduce lifespan of a component is the overclock producing higher heat. Nothing more, nothing less. However if you can keep the average temperature of the GPU the same as what it was expected to run at, there won't be a significant difference in life span.

[Enderman]

4 hours ago, fasauceome said:

Ah ok I'll just set my Vcore to 5 volts then, I'm sure it's fine.

 

Really though, semiconductors wear out due to the kinetic energy of electrons degrading the internal transistors, even if you kept a CPU at 30 degrees it wouldn't last forever. A copious amount of voltage at a safe temperature will cause problems and a safe voltage at a copious temperature will cause problems. It's not the opposite of the truth, it's only half of what you need to look out for.

 

And yes, I know what popcorning is.

If you set vcore to 5V it will only damage the transistors because of heat.

If you kept it at 30C it WOULD last just as long as a regular CPU at regular voltages.

The problem is you CAN'T keep it at 30C because you can't cool the transistors directly, only the top layer of silicon of the CPU.

The distance between the area of cooling and the transistor is too far, causing a huge temperature gradient.

Even though the surface of the CPU might be at -272C the transistors would be far above 30C and would die instantly.

 

Again, it is the temperature that causes the damage, not the voltage.

 

If you have a transistor that isn't buried under layers of silicon you can do 7V+ and even more if you cool it actively.

It's all about how how the actual transistor gets.

[ewitte]

The cards are locked from using dangerous voltages.  With the amount of control you have it will still outlive its usefulness.

[Enderman]

4 hours ago, Kalm_Traveler1 said:

in the context of this conversation, voltage will indeed kill things. No need to purposely mislead someone who is trying to learn

In the context of this conversation, it is heat, caused by increasing the voltage, which can kill things.

The temperature also depends on OP's cooling method, not the voltage alone.

If he has a good cooler on it, he can easily increase his voltage by a few hundred millivolts and the temperatures will still be fine, and his CPU/GPU lifespan will not be reduced.

 

No need to mislead people and tell them that more voltage will kill their CPU/GPU, when it will not.

More heat will.

That is why you need to monitor temperatures under load when overclocking, and buy a new cooler if necessary.

 

In fact, nvidia GPUs are limited at a certain max power and offset, so you can give them all the voltage and power you want and it will be limited to safe amounts anyway.

 

[Fasauceome]

23 minutes ago, Enderman said:

If you set vcore to 5V it will only damage the transistors because of heat.

If you kept it at 30C it WOULD last just as long as a regular CPU at regular voltages.

The problem is you CAN'T keep it at 30C because you can't cool the transistors directly, only the top layer of silicon of the CPU.

The distance between the area of cooling and the transistor is too far, causing a huge temperature gradient.

Even though the surface of the CPU might be at -272C the transistors would be far above 30C and would die instantly.

 

Again, it is the temperature that causes the damage, not the voltage.

 

If you have a transistor that isn't buried under layers of silicon you can do 7V+ and even more if you cool it actively.

It's all about how how the actual transistor gets.

Actually Hardcore Overclocking describes voltage and heat damage as two separate characteristics

 

Steve burke covered this, and it's a well understood phenomena

 

https://www.edn.com/electronics-news/4382044/What-Causes-Semiconductor-Devices-to-Fail-

different types of failure

http://www.circuitstoday.com/pn-junction-breakdown-characteristics

junction breakdown

 

edit:

check out this video at 9:30 in

Quote

either indirectly from too much heat, or directly from just being too much voltage

 

[Kalm_Traveler]

2 minutes ago, Enderman said:

In the context of this conversation, it is heat, caused by increasing the voltage, which can kill things.

The temperature also depends on OP's cooling method, not the voltage alone.

If he has a good cooler on it, he can easily increase his voltage by a few hundred millivolts and the temperatures will still be fine, and his CPU/GPU lifespan will not be reduced.

 

No need to mislead people and tell them that more voltage will kill their CPU/GPU, when it will not.

More heat will.

That is why you need to monitor temperatures under load when overclocking, and buy a new cooler if necessary.

 

In fact, nvidia GPUs are limited at a certain max power and offset, so you can give them all the voltage and power you want and it will be limited to safe amounts anyway.

 

You just reiterated my point, so thank you.

 

For someone new to this type of scene, telling them not to worry about voltage because "only heat kills" is a terrible idea. 

 

Too much voltage will indeed kill, both directly and indirectly as you will learn in any basic electronics class.

[Enderman]

9 minutes ago, fasauceome said:

Actually Hardcore Overclocking describes voltage and heat damage as two separate characteristics

Steve burke covered this, and it's a well understood phenomena

 

https://www.edn.com/electronics-news/4382044/What-Causes-Semiconductor-Devices-to-Fail-

different types of failure

http://www.circuitstoday.com/pn-junction-breakdown-characteristics

junction breakdown

1) If you read your first link, which I assume you're talking about electrical overstress, everything described there about derating components and electrical stress has to do with heat. The damage is caused by heat (except in the rare case where it arcs and shorts, which is caused by ESD)

All the other stuff has to do with heat.

All those transistors and diodes have a resistance that causes an increase in current when you add voltage. This in turn heats up the components and traces and everything else.

 

2) says the same thing I've been telling you repeatedly.

It allows lots of current through and causes damage.

Guess why? Because heat.

 

[Mick Naughty]

Saying it all like this guy is gonna do much more than move the slider up on afterburner. 

Which in turn won’t do anything. Unless the card is supposed to last 10 years and dies in 8. I’m sure he will be just fine.