F5F Stay Refreshed Power Users Overclocking Is it feasible to keep overclocking at 6GHz or higher for a long time?

Is it feasible to keep overclocking at 6GHz or higher for a long time?

Is it feasible to keep overclocking at 6GHz or higher for a long time?

V
Viridian
Member
168
09-09-2018, 12:01 PM
#1
So I was checking geekbench 4 and seeing the top scores for singlecore and multicore. They're really high. I have a few questions.

1.) The Intel i9-9900k runs between 3.6 and 5.0 GHz. If I kept spraying liquid nitrogen on it to reach 6-7GHz, would it still stay at that speed? Or would it drop back down to 3.6-5GHz eventually? If I could maintain 6-7GHz without problems, would it actually help and notice a boost in performance?
2.) For the Intel Xeon Platinum 9282 and AMD Epyc 7742, would adding liquid nitrogen work? And why wouldn’t big companies like Amazon Web Services do this? Or would it not be practical or worth it?
3.) Does running the same processor at 3.6GHz versus 5GHz affect game speed in a game like Minecraft?
V
Viridian
09-09-2018, 12:01 PM #1

So I was checking geekbench 4 and seeing the top scores for singlecore and multicore. They're really high. I have a few questions.

1.) The Intel i9-9900k runs between 3.6 and 5.0 GHz. If I kept spraying liquid nitrogen on it to reach 6-7GHz, would it still stay at that speed? Or would it drop back down to 3.6-5GHz eventually? If I could maintain 6-7GHz without problems, would it actually help and notice a boost in performance?
2.) For the Intel Xeon Platinum 9282 and AMD Epyc 7742, would adding liquid nitrogen work? And why wouldn’t big companies like Amazon Web Services do this? Or would it not be practical or worth it?
3.) Does running the same processor at 3.6GHz versus 5GHz affect game speed in a game like Minecraft?

M
ManicFG
Member
72
09-20-2018, 09:44 AM
#2
Certainly understood. Here is the rewritten version:

1: Of course you could, but I question whether you'd be able to locate a dependable method without disrupting the rest of the board; insulation only provides limited effectiveness.
Beneficial? Not really.
2:
It offers only a small improvement in performance for a huge investment and the danger of running LN2. It's not worth it to spend money on a poorly fitting and ineffective cooling system when you could simply purchase an extra processor.
3: The answer varies depending on the processor.
In summary, if LN2 were actually feasible, it would remain outside the realm of overclocking achievements.
M
ManicFG
09-20-2018, 09:44 AM #2

Certainly understood. Here is the rewritten version:

1: Of course you could, but I question whether you'd be able to locate a dependable method without disrupting the rest of the board; insulation only provides limited effectiveness.
Beneficial? Not really.
2:
It offers only a small improvement in performance for a huge investment and the danger of running LN2. It's not worth it to spend money on a poorly fitting and ineffective cooling system when you could simply purchase an extra processor.
3: The answer varies depending on the processor.
In summary, if LN2 were actually feasible, it would remain outside the realm of overclocking achievements.

S
Shanniee_
Member
104
09-20-2018, 11:39 AM
#3
Certainly understood. Here is the rewritten version:

1: Of course you could, but I question whether you'd be able to locate a dependable method without disrupting the rest of the board; insulation only provides limited effectiveness.
Beneficial? Not really.
2:
A very small improvement, but it's hardly worthwhile. A huge investment and a dangerous run of LN2—why invest in a poorly designed cooling system when you could simply purchase more processors?
3: It hinges on the processor type.
In summary, if LN2 were ever feasible, it would remain outside the realm of overclocking achievements.
S
Shanniee_
09-20-2018, 11:39 AM #3

Certainly understood. Here is the rewritten version:

1: Of course you could, but I question whether you'd be able to locate a dependable method without disrupting the rest of the board; insulation only provides limited effectiveness.
Beneficial? Not really.
2:
A very small improvement, but it's hardly worthwhile. A huge investment and a dangerous run of LN2—why invest in a poorly designed cooling system when you could simply purchase more processors?
3: It hinges on the processor type.
In summary, if LN2 were ever feasible, it would remain outside the realm of overclocking achievements.