goal of overclocking?

[jimmyjimmies]

Is the goal to get the highest GHz frequency with the lowest temps(due to voltage right?) as possible. so if I get 4.7 GHz frequency at a 80 or so degress, I should keep lowering voltage until the stress test fails and then go back to previous voltage? What happens if you run too high of frequency without enough voltage? slow apps? apps force close? blue screen?

 

thanks

[The Sloth]

depends on the chip

[jimmyjimmies]

Just now, nerdslayer1 said:

depends on the chip

7700k

[The Sloth]

Just now, jimmyjimmies said:

7700k

most 7700k gets about 4.9 ghz, some even goes high as 5.0 to 5.3 ghz so it depends, don't go over 1.3v for daily usage, no matter what you do do not go over 1.4v. 

[jasonc_01]

WORLD DOMINATION LOL

[jimmyjimmies]

1 minute ago, nerdslayer1 said:

most 7700k gets about 4.9 ghz, some even goes high as 5.0 to 5.3 ghz so it depends, don't go over 1.3v for daily usage, no matter what you do do not go over 1.4v. 

what im asking is if my temps are too high, then should i lower my voltage at same frequency? and what happens when you do that and your voltage is too low?

[The Sloth]

Just now, jimmyjimmies said:

what im asking is if my temps are too high, then should i lower my voltage at same frequency? and what happens when you do that and your voltage is too low?

 

 

[jimmyjimmies]

1 minute ago, nerdslayer1 said:

 

seen it

[SpaceGhostC2C]

4 minutes ago, jimmyjimmies said:

what im asking is if my temps are too high, then should i lower my voltage at same frequency? and what happens when you do that and your voltage is too low?

Random crashes, BSOD, or even failing to POST.

[The Sloth]

2 minutes ago, jimmyjimmies said:

seen it

follow that

[DildorTheDecent]

This is the goal:

[jimmyjimmies]

4 minutes ago, SpaceGhostC2C said:

Random crashes, BSOD, or even failing to POST.

thank you for answering the question

[jimmyjimmies]

Just now, DildorTheDecent said:

This is the goal:

lol i wish, looks like im stuck until i figure out the cooling though

[The Sloth]

Just now, Aleksiandrovich said:

Umm not true, thats a misconception, in fact i believe that Sky/KabyLake has higher voltage maximums than the previous generations. SkyLake can safely be at 1.4V for daily use, as can be Kabylake - @Lays for reference.

 

...no it's not recommended going over 1.39v 

[DildorTheDecent]

Just now, jimmyjimmies said:

lol i wish, looks like im stuck until i figure out the cooling though

Get a good 280 or 360 AIO. 7700K needs it. 

 

If you can afford a custom water loop do that instead. You're good for 1.35V - 1.4V+ on custom water. That'll get you 5.0GHz+ at least. 

[The Sloth]

1 minute ago, Aleksiandrovich said:

As i said, the SkyLake chip that Lays had ran at 1.46V for daily use on a loop, with KabyLake having higher tolerance.

you can easily go 4.9ghz or 5 ghz depending on chip without going past 1.39v.

[DildorTheDecent]

2 minutes ago, Aleksiandrovich said:

As i said, the SkyLake chip that Lays had ran at 1.46V for daily use on a loop, with KabyLake having higher tolerance.

Binned chips and delid is what made that possible. And a massive 1080mm radiator. 

 

Still awesome though. OCN is the better place to find proper overclocking. Plenty of guys push the volts for the clock there. 

 

Heck, the whitepaper specs 1.52V as max. Plenty of room. 

[Pidgey]

 

11 minutes ago, DildorTheDecent said:

Get a good 280 or 360 AIO. 7700K needs it. 

 

If you can afford a custom water loop do that instead. You're good for 1.35V - 1.4V+ on custom water. That'll get you 5.0GHz+ at least. 

 

9 minutes ago, Aleksiandrovich said:

As i said, the SkyLake chip that Lays had ran at 1.46V for daily use on a loop, with KabyLake having higher tolerance.

could the reason im hitting 89 degrees at 1.325 volts at 4.8 on 15 min realbench stress test be because im using 240mm rad instead of 280? and when i say hit 89, i mean that was the max that a core hit

[The Sloth]

1 minute ago, Aleksiandrovich said:

As i said, i didn't specify what could be or couldn't be achieved on what voltage but rather what voltage is OK before it becomes unsafe, if you want to use a higher voltage such as low to mid 1.4 then by all means, it is still a safe voltage. As @DildorTheDecent said the paper specs are 1.52V max (Yes i know the circumstances of @Lays system are very specific but still it was a high voltage still being safe for use, as in not destroying the CPU. Might i add that if the temperatures are fair comparative to the voltage then it is fine to push it, 1.39V is not max for the CPU and as such if you have temperature head room then by all means you can safely push the voltage higher.)

 

 

really don't care what @DildorTheDecent says 1.52v is absurd for a 7700k, OP is a beginner so starting him with delidding a cpu is not the best idea. going 1.3v and trying his luck with overclocking is a safe bet for OP than trying higher overclocks in the future. 

[DildorTheDecent]

5 minutes ago, Pidgey said:

 

 

could the reason im hitting 89 degrees at 1.325 volts at 4.8 on 15 min realbench stress test be because im using 240mm rad instead of 280? and when i say hit 89, i mean that was the max that a core hit

Really it's down to realbench. It's the better CPU stress utility out there. Really goes hard on a CPU with a realistic workload. And as such really kicks out heat. 

 

Kaby Lake from the limited Googling just appears to run a little warm. Lower process and all those transistors will run a bit toasty. 

 

If anything, a delid will get you some proper good thermal headroom. Drops of 20C are very possible. 240mm is still a good rad choice, but the way pricing is, a 280 could probably have worked better. But sure. Get a delid and temps will come down. 

4 minutes ago, Aleksiandrovich said:

As i said, i didn't specify what could be or couldn't be achieved on what voltage but rather what voltage is OK before it becomes unsafe, if you want to use a higher voltage such as low to mid 1.4 then by all means, it is still a safe voltage. As @DildorTheDecent said the paper specs are 1.52V max (Yes i know the circumstances of @Lays system are very specific but still it was a high voltage still being safe for use, as in not destroying the CPU. Might i add that if the temperatures are fair comparative to the voltage then it is fine to push it, 1.39V is not max for the CPU and as such if you have temperature head room then by all means you can safely push the voltage higher.)

I believe since Ivy Bridge Intel just threw down 1.52V as the max as that's all anyone would ever need. They can be run much higher with exotic or extreme cooling but that's the exceptional circumstance. 

 

Just now, nerdslayer1 said:

really don't care what @DildorTheDecent says 1.52v is absurd for a 7700k, OP is a beginner so starting him with delidding a cpu is not the best idea. going 1.3v and trying his luck with overclocking is a safe bet for OP than trying higher overclocks in the future. 

Thanks man. Always knew you supported me. <3 

 

Shame that "enthusiast computing" is completely gone on you though :c

[Pidgey]

okay thanks guys!

[The Sloth]

2 hours ago, DildorTheDecent said:

Really it's down to realbench. It's the better CPU stress utility out there. Really goes hard on a CPU with a realistic workload. And as such really kicks out heat. 

 

Kaby Lake from the limited Googling just appears to run a little warm. Lower process and all those transistors will run a bit toasty. 

 

If anything, a delid will get you some proper good thermal headroom. Drops of 20C are very possible. 240mm is still a good rad choice, but the way pricing is, a 280 could probably have worked better. But sure. Get a delid and temps will come down. 

I believe since Ivy Bridge Intel just threw down 1.52V as the max as that's all anyone would ever need. They can be run much higher with exotic or extreme cooling but that's the exceptional circumstance. 

 

Thanks man. Always knew you supported me. <3 

 

Shame that "enthusiast computing" is completely gone on you though :c

 

 

you keep reminding me why i blocked you. 

[Lays]

11 hours ago, nerdslayer1 said:

...no it's not recommended going over 1.39v 

And where did you see this?
 

If going over 1.39v was dangerous, 1.39v would be dangerous too.  You clearly have no idea or personal experience about what you're giving advice on.  With proper cooling & a delid, 1.4-1.425v is easily coolable under reasonable gaming/render loads.  The general rule of thumb of "stop when you start hitting 80's in stress tests" applies here.

Previously to buying my 7700k, I was dailying a 6700k at 1.43v bios at 5ghz for nearly a year before I sold it. I had gone to 1.7v+ on dry ice multiple times, and saw no issues with degradation. It still ran 5ghz daily at the same voltage.  There's countless threads on OCN and other forums where hardware reps, and people that have lots of overclocking experience using 1.4v (and higher) with no problems.

 

Temperatures have a WAY bigger impact in my experience vs voltage, if you try running 1.6v through your chip for a few hours on water, you will see degradation, do the same thing on dry ice and it's no problem.  It's all about temperature vs voltage.  The higher the voltage, the cooler you should try to keep the chip to prevent that extra voltage from damaging things.

[Goat]

While the goal is achieve the highest frequency possible, there are factors that prevent you from running the CPU at your desired speed, forcing you to reduce the overclock:

 

If you run at a high frequency without enough voltage to make the chip stable, you will get BSOD. That is why voltage is increased to keep the overclock stable, at the cost of temperatures. If temps are considered high for the current voltage that you are running ( 80c and above ), reduce the voltage to keep temps in check and dial down over overclock a bit to keep things stable. Another option would be to get a better cooler. 

 

As one of the main killers of PC components is heat, I would refrain from constantly running the CPU at 80c or above. Running at 80c is not considered as overheating ( yet ), but it is considered hot.

 

Another factor that is out of your control is the "silicon lottery". Its basically a term to describe how "good" a chip is at overclocking, as not all chips, even ones of the same model, are created equally due to the manufacturing process. Imagine two CPUs of the exact same model running the same cooler etc. CPU A can overclock up to 4.8ghz while CPU B can only overclock up to 4.5ghz, where it will crash if you try to push it further. This is an example of the silicon lottery in action, where even though you have two CPUs of the same model, one cannot overclock as well as the other due to manufacturing. If by chance you get a "CPU A" that can overclock well, then you have "won" the silicon lottery. Good for you. But, if you happen to get a "CPU B" that can't overclock well, then you simply won't be able to achieve the same overclocks that other people are getting with the same chip without crashing.

 

So with these factors in mind, you will want to overclock your CPU as far as it can go while keeping things as cool and stable as possible. You can use other people's overclocks as reference, but note that yours will either be higher or lower than theirs due to the silicon lottery.

 

 

[Pidgey]

30 minutes ago, Goat said:

While the goal is achieve the highest frequency possible, there are factors that prevent you from running the CPU at your desired speed, forcing you to reduce the overclock:

 

If you run at a high frequency without enough voltage to make the chip stable, you will get BSOD. That is why voltage is increased to keep the overclock stable, at the cost of temperatures. If temps are considered high for the current voltage that you are running ( 80c and above ), reduce the voltage to keep temps in check and dial down over overclock a bit to keep things stable. Another option would be to get a better cooler. 

 

As one of the main killers of PC components is heat, I would refrain from constantly running the CPU at 80c or above. Running at 80c is not considered as overheating ( yet ), but it is considered hot.

 

Another factor that is out of your control is the "silicon lottery". Its basically a term to describe how "good" a chip is at overclocking, as not all chips, even ones of the same model, are created equally due to the manufacturing process. Imagine two CPUs of the exact same model running the same cooler etc. CPU A can overclock up to 4.8ghz while CPU B can only overclock up till 4.5ghz, where it will crash if you try to push it further. This is an example of the silicon lottery in action, where even though you have two CPUs of the same model, one cannot overclock as well as the other due to manufacturing. If by chance you get a "CPU A" that can overclock well, then you have "won" the silicon lottery. Good for you. But, if you happen to get a "CPU B" that can't overclock well, then you simply won't be able to achieve the same overclocks that other people are getting with the same chip without crashing.

 

So with these factors in mind, you will want to overclock your CPU as far as it can go while keeping things as cool and stable as possible. You can use other people's overclocks as reference, but note that yours will either be higher or lower than theirs due to the silicon lottery.

 

 

Okay, thanks for the info, I just wanted to confirm those things to diagnose my problem. BC there's no reason for me to be hitting over 80 celcius at 1.25 volts on a 15 minutes real bench stress test.Do you know if some chips may have worse TIM application from the factory, or just have something defective to cause higher temps? I've reapplied paste, checked cooler, fans, air circulation, its all new and functional. kraken x52 is a 240mm AIO, but its one of the best, and is capable of much better results. So I'm starting to think it might be an issue with the chip?