Variance in Chips (Overclocking)

[Pidgey]

Where is the actual variance in chips? Here's what I've gathered reading through various posts and peoples confusing arguments on this subject. I am new to overclocking, so please let me know what I've misinterpreted and provide clarification if possible. Thanks!

-All chips of the same kind(i.e. kaby lake) can be set to 5.0 ghz, 5.2 ghz, or any speed, but they vary by what voltage is required to maintain that clock speed.
-Clock speed has minor(if any) direct impact on temperature, but indirectly effects temps through required voltage increases.
-Testing multiple chips of the same kind in the same cooling environment/build would result in the same, or very similar temperatures at any given voltage.

-So the only major variance in chips is the voltage required to meet a certain clock speed.

 

Hoping to learn more about overclocking and narrow down the possible causes of my high cpu temps. If everything above is true, then I'll assume it's an issue with my my AIO.

 

Here's the issue for anyone curious-- I ran a 15 minute stress test with clock speed of 4.8 ghz at 1.275 volts, The max temp reached 83 Celsius. 4.8 at 1.275 would leads me to assume I could probably reach 5.0+ pushing 1.3-1.4 voltage, but my temps went to 90 when I stress tested at 1.3.

 

i7-7700k

kraken x52 AIO

maximus ix formula

9 fans- 4 top/rear exhaust, 5 front intake push/pulling rad

[SpaceGhostC2C]

32 minutes ago, Pidgey said:

 

-All chips of the same kind(i.e. kaby lake) can be set to 5.0 ghz, 5.2 ghz, or any speed, but they vary by what voltage is required to maintain that clock speed.

Not sure all chips can do it at some voltage. But it could be that "some voltage" means 2.35v or something like that for some chips, which is the equivalent of not possible  But you can certainly have a chip that can't do what other chips achieve at any voltage that won't burn your VRM in seconds.

32 minutes ago, Pidgey said:

-Clock speed has minor(if any) direct impact on temperature, but indirectly effects temps through required voltage increases.

Yes, that's true (I lean on the "minor" side rather than "none").

32 minutes ago, Pidgey said:

-Testing multiple chips of the same kind in the same cooling environment/build would result in the same, or very similar temperatures at any given voltage.

I would expect so.

32 minutes ago, Pidgey said:

-So the only major variance in chips is the voltage required to meet a certain clock speed.

Yep (and possibly whether certain clocks are within reach at all).

[Pidgey]

1 minute ago, SpaceGhostC2C said:

Not sure all chips can do it at some voltage. But it could be that "some voltage" means 2.35v or something like that for some chips, which is the equivalent of not possible  But you can certainly have a chip that can't do what other chips achieve at any voltage that won't burn your VRM in seconds.

Yes, that's true (I lean on the "minor" side rather than "none").

I would expect so.

Yep (and possibly whether certain clocks are within reach at all).

thanks, didn't actually expect to be right on most of those lol! so would you agree that the issue has to be with my cooling, or more specifically my aio?

[SpaceGhostC2C]

1 minute ago, Pidgey said:

thanks, didn't actually expect to be right on most of those lol! so would you agree that the issue has to be with my cooling, or more specifically my aio?

I'm not too familiar with 7700K overclocking, so I'm afraid I can't tell you how hot you can expect it to get at those voltages.

Which test are you running? Temperatures also vary from test to test, since they perform different tasks with the CPU.

[Pidgey]

Just now, SpaceGhostC2C said:

I'm not too familiar with 7700K overclocking, so I'm afraid I can't tell you how hot you can expect it to get at those voltages.

Which test are you running? Temperatures also vary from test to test, since they perform different tasks with the CPU.

I'm running real bench stress test, with 8-16gb ram, not sure if the ram tested makes a difference though