Overclock.net banner

7700K at 5.4GHz

4K views 26 replies 8 participants last post by  JCOC 
#1 ·
I was able to boot and validate at 5.4GHz. After I make some adjustments, I will update this thread with some results from a few benchmarks.



CPU-Z Validation Link:

https://valid.x86.fr/p3d6pq

System configuration:
Z270-S-7700K

I delidded it this time. Which is the first time I have overclocked with it delidded. I used Thermal Grizzly Conductonaut (liquid metal) under the IHS. I used Artic Silver 5 between the water block and CPU.

I have four 140mm Corair SP140 fans configured in a push/pull setup on the rad of the H115i.

The average RPM speed on the pump was 3000 RPMS

I was running the stock frequency of 2133MHzCL15 on my RAM.

thinking.gif
Seriously wIth an AIO? Whats wrong with you?
hmmsmiley02.gif

eh-smiley.gif
Ehh I know, but I guess it is what it is.
eh-smiley.gif
 
See less See more
6
#3 ·
Quote:
Originally Posted by czin125 View Post

Corsair H115i with 2x default fans at max rpm?

Delidded? Liquid metal?

What did you use for IHS -> Waterblock?

79C max with 2667CL15 running.
I'm sorry for not explaining that information in the original thread. I will update it now.

I delidded it this time. Which is the first time I have overclocked with it delidded. I used Thermal Grizzly Conductonaut (liquid metal) under the IHS. I used Artic Silver 5 between the water block and CPU.

I have four 140mm Corair SP140 fans configured in a push/pull setup on the rad of the H115i.

The average RPM speed on the pump was 3000 RPMS

I was running the stock frequency of 2133MHzCL15 on my RAM.
 
#4 ·
Nice. Got my delid kit coming in the mail, should be here Saturday. Liquid metal won't be here until sometime next week though.

Run Super Pi
smile.gif
 
#7 ·
Rockit 88. Had to order the CLU separate since it was OOS so it won't be until next week when it's all done unfortunately.
frown.gif
 
#11 ·
Quote:
Originally Posted by czin125 View Post

A delid should allow you to drop the voltage due to the -25-30C drop. Are you also getting Kryonaut for the IHS->Waterblock?
That is where most people get confused. You can not drop voltage just because your temps drop. Every CPU that runs at "X" frequency will have to use "X" amount of voltage in or to operate.

Lets say this-
I am running 5.2GHz at 1.34V with temps around 85°C-90°C under load. I noticed that about 3 minutes into my benchmark, I am thermal throttling because 85°C-90°C are too high. Now I decide to you delid my CPU. The only thing that will change is the temps. I will still have to use 1.34V to get 5.2GHz, however; my CPU will no longer thermal throttle and my benchmark score will be higher.
 
#13 ·
Quote:
Originally Posted by czin125 View Post

Are you also getting Kryonaut for the IHS->Waterblock?
I have purchased Kryonaut 3 times and all three times the pasted dried up in a month. Granted, I was hitting 100°C almost every day. Lol I just cant bring myself to buy another tube of Kryonaut. Artic Silver 5 is where its at!! I've been rocking one application for 3 months now. So in my opinion Kryonaut is over rated
 
#16 ·
We all OC differently. We all want different results. I dont care for a daily use OC. I want the fastest frequency that is benchmarkable. Look at the table (under the 1151 socket tab) in the 5GHz Overclock Club. "The lucky" chips hit 5.4GHZ at 1.4~ volts. The unlucky chips have to use 1.4~ volts to hit 5.2GHz. Temps did not hold the "lucky" or "unlucky"' chips back. The voltage did. It's all about how your system cools the amount of voltage applied.
 
#17 ·
Increases in temperature will increase the resistance of a wire (i.e. where your voltage is flowing in the CPU), then taking Ohm's law... V = IR, V is proportional to the resistance, so R changes, Voltage does to. This means that temperature affects voltage. I saw this effect going from air+lidded to custom water+delidded pretty nicely on my 4790k (iirc dropped ~0.05v for the same clocks).

Back on topic, thats a very nice OC
thumb.gif
 
#18 ·
Its all about how your systems cools the amount of voltage applied

I don't how to explain it any better. Now if you would like to put it in different words to feel better. Go right ahead. Delid if you want lower temps in your daily use OC. Or in my case, delid if you want to eliminate as much thermal throttling as possible. Every chip is different. Many have different cooling environments and ambient temperatures. It is a lie when anyone guarantees or promises to anyone that delidding will FOR SURE lower your amount of voltage needed for the same clock.
 
#19 ·
Quote:
Originally Posted by JCOC View Post

Its all about how your systems cools the amount of voltage applied
Could you use a larger text? Sorry, my eye sight isn't the best.

biggrin.gif
 
  • Rep+
Reactions: Koniakki
#21 ·
Quote:
Originally Posted by The Pook View Post

My old Sempron 145 wouldn't post above 4Ghz even with 1.75v. Put it under DICE and it would. Not relal
Could you use a larger text? Sorry, my eye sight isn't the best.

biggrin.gif
I apologize for my inconsideration. Please allow me to correct myself.

It's all about how your system cools the amount of voltage applied
 
#22 ·
JCOC
For years it's been that lower temperatures have reduced voltage requirements. Also this is too broad of a subject as well. for example, CPU X at 5 ghz may require 1.35v for stability. But when temps exceed 80C (NOT Throttling), prime95 will error out. So what do you do?
you raise the voltage to 1.38v.

Now Prime95 runs just fine, but now temps go up to 90C instead of 80C.

So, you delid, temps drop to 65C maximum at 1.35v. Now prime95 runs 8 hours and doesn't drop a core, because it never reaches the temp where things get unstable due to heat and resistance (80C).

You may be able to even reduce vcore a few notches and maintain stability AND lower temps, but you would have to be careful, because, go too low, you may find that Prime will drop a core at 5 ghz at 1.30v at 60C.

This is also how LN2 works. Even though no one uses LN2 for normal clocks. Cool that 5 ghz CPU down to -50C, it may be stable at 5 ghz at *1.20v*.
 
#23 ·
Quote:
Originally Posted by Falkentyne View Post

JCOC
For years it's been that lower temperatures have reduced voltage requirements. Also this is too broad of a subject as well. for example, CPU X at 5 ghz may require 1.35v for stability. But when temps exceed 80C (NOT Throttling), prime95 will error out. So what do you do?
you raise the voltage to 1.38v.

Now Prime95 runs just fine, but now temps go up to 90C instead of 80C.

So, you delid, temps drop to 65C maximum at 1.35v. Now prime95 runs 8 hours and doesn't drop a core, because it never reaches the temp where things get unstable due to heat and resistance (80C).

You may be able to even reduce vcore a few notches and maintain stability AND lower temps, but you would have to be careful, because, go too low, you may find that Prime will drop a core at 5 ghz at 1.30v at 60C.

This is also how LN2 works. Even though no one uses LN2 for normal clocks. Cool that 5 ghz CPU down to -50C, it may be stable at 5 ghz at *1.20v*.
Exactly. But what do I know, only have an EE MS and work in the semi industry...
 
#25 ·
Quote:
Originally Posted by asuindasun View Post

Exactly. But what do I know, only have an EE MS and work in the semi industry...
Well I have an Auto Body Technician Diploma ... checkmate.
 
#26 ·
Quote:
Originally Posted by asuindasun View Post

Exactly. But what do I know, only have an EE MS and work in the semi industry...
Thats cool but it still doesn't change the fact you can't guarantee lower voltage on the same clock just because you deledded the CPU. Which was my point in the first place. Do you have a higher chance of lowering the voltage, yes. It's not a guarantee though.@Falkentyne said so him self when said "You may be able to even reduce the vcore."

KEYWORD: may
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top