Delidding Intels 7700k drops temps by almost 30C

[MorganO]

8 hours ago, brighttail said:

Thankfully Intel has some more time to deal with their thermal issues.

 

When originally tested, the 7700k CPU could obtain 5Ghz speed with 1.35v but was hitting temps of 100C!

"By delidding the CPU and applying CoolLabatory Liquid Thermal Paste and a Kraken x62 cooler, they were able to push it to 5.0Ghz using 1.344v with the fan set to 50%. The pump was at 65% and yet the temps plummeted by 30C.  An average of 26C."

 

Intel is set to launch Kaby Lake with the I7-7700k at CES in January 2017 so they have a little time to work on the thermals.

 

http://www.tweaktown.com/news/55497/7700k-delidded-30c-reduction-temps-wtf-intel/index.html

 

I looked to see if i missed this post already out there, I probably missed it 10 pages back so sorry if it is a repeat!

 

NOTE ** This is not a recommendation to delid ANY CPU!  It is advanced knowledge and if you have any questions I suggest you google the video Linus did on the procedure.  The point of this post was to show that the thermal compound Intel is using on these pre-production and review CPUs MAY result in much higher temperatures.

please edit this and point out this 

New work has been done on the 7700K with retail samples, and cooled by the Thermalright 6 cooler by AnandTech forum member 'RichUK'. Out of the box, the temperatures of the Core i7-7700K hit 60C under load, but when the 7700K was pushed to 5GHz and pushed to 100% load, the temperatures spiked at a mammoth 96C" 

 

[leadeater]

@KOMTechAndGaming

 

At least give a person long enough to see your post before posting the same information multiple times, particularly when there hasn't even been any more posts in the thread after yours so thinking it could be missed/not read by people shouldn't be a factor.

[Coaxialgamer]

12 hours ago, Notional said:

Or you could wait for ZEN, which are soldered, instead of paying overprices for few cores and shitty thermal paste?

Has AMD said anything about the die being soldered? 

Intel isn't using TIM on their cpu's JUST because they are greedy . The dies are getting so small and thin that there isn't enough room to offset the strain caused by the application and use of solder on the die.

[Coaxialgamer]

8 hours ago, AluminiumTech said:

I thought we gave up on the GHz race over 10 years ago.........

 

Why do we need a 5GHz CPU? We need a 3-4GHz CPU with significantly better IPC. Not some kind of BS overclocking for better performance.

i kind of agree with you , but unfortunately we aren't going to see any huge IPC improvements on desktop , not on x86 and not with how current software works anyway.

And that is regardless of how much effort intel puts into their architecture development. 

[Sauron]

Sandy Bridge remains the best launch intel did in the last 20 years.

[MorganO]

2 hours ago, leadeater said:

@KOMTechAndGaming

 

At least give a person long enough to see your post before posting the same information multiple times, particularly when there hasn't even been any more posts in the thread after yours so thinking it could be missed/not read by people shouldn't be a factor.

sorry my bad

Edited December 16, 2016 by KOMTechAndGaming
removed caps

[TheRandomness]

13 hours ago, Notional said:

Or you could wait for ZEN, which are soldered, instead of paying overprices for few cores and shitty thermal paste?

Don't you mean wait for benchmarks?

[Notional]

2 hours ago, Coaxialgamer said:

Has AMD said anything about the die being soldered? 

Intel isn't using TIM on their cpu's JUST because they are greedy . The dies are getting so small and thin that there isn't enough room to offset the strain caused by the application and use of solder on the die.

 

 

Ryzen SR7, or XEON; depends on who you are.

40 minutes ago, TheRandomness said:

Don't you mean wait for benchmarks?

Benchmarks doesn't say anything about pricing, which is equally as important. We have seen benchmarks enough to get an idea of performance though.

[Coaxialgamer]

2 hours ago, Notional said:

 

Ryzen SR7, or XEON; depends on who you are.

Benchmarks doesn't say anything about pricing, which is equally as important. We have seen benchmarks enough to get an idea of performance though.

That most likely the 16 core model. 

There's no guarantee the 4-8 cores will use solder,  considering they will be smaller. 

Plus,  when reading that rumor,  it appeared fairly likely ( as the author of the article also supposed)  that thus was in fact a bristol ridge part ( 28nm, excavator) 

[Coaxialgamer]

4 hours ago, Sauron said:

Sandy Bridge remains the best launch intel did in the last 20 years.

Pentium M (banias,  dothan) was pretty fucking great back in it's day,  but it didn't get much attention because it was a mobile processor.  But it did lay much of the ground work for the core series. 

[Notional]

21 minutes ago, Coaxialgamer said:

That most likely the 16 core model. 

There's no guarantee the 4-8 cores will use solder,  considering they will be smaller. 

Plus,  when reading that rumor,  it appeared fairly likely ( as the author of the article also supposed)  that thus was in fact a bristol ridge part ( 28nm, excavator) 

 

That is an 8 core (2xquad core). The bigger chips are not FM4, but that giant LGA platform.

[Coaxialgamer]

2 hours ago, Notional said:

That is an 8 core (2xquad core). The bigger chips are not FM4, but that giant LGA platform.

If that is actually Ryzen SR7 , knowing that the am4 package size is 40mm*40mm , that would put this die at about ~250mm² . 

[Notional]

5 minutes ago, Coaxialgamer said:

If that is actually Ryzen SR7 , knowing that the am4 package size is 40mm*40mm , that would put this die at about ~250mm² . 

Do we know it's 40x40mm?

 

A 10 core Broadwell-e is a tiny bit smaller. But all that north and south bridge stuff implemented into zen will take up space.

[Coaxialgamer]

5 minutes ago, Notional said:

Do we know it's 40x40mm?

 

A 10 core Broadwell-e is a tiny bit smaller. But all that north and south bridge stuff implemented into zen will take up space.

just looked up " am4 package size" on google . From leaks and patents it appears to be 40*40mm , which seems reasonable. Couldn't find any more info though . It should also be noted that paint ( which i used ) isn't the most accurate of tools, so i could be off by +/- 10% .

 

I do find it strange that the SR7 die would be split into two parts like that though , i've seen nothing pointing to that.

 

It's worth noting that bristol ridge apu's actually clock in at ~240mm² , which isn't far off from this. 

I wouldn't be so sure this HAS to be Zen

[Notional]

16 minutes ago, Coaxialgamer said:

It's worth noting that bristol ridge apu's actually clock in at ~240mm² , which isn't far off from this. 

I wouldn't be so sure this HAS to be Zen

 

 

In case it IS Bristol Ridge, and not Summit Ridge, I think that makes it even more realistic to be soldered, as BR SR is a more expensive part.

[Coaxialgamer]

10 minutes ago, Notional said:

In case it IS Bristol Ridge, and not Summit Ridge, I think that makes it even more realistic to be soldered, as BR is a more expensive part.

well bristol ridge IS bigger , but not necessarily more expensive . Wafer costs have gone way up since 28nm

[patrickjp93]

11 minutes ago, Notional said:

In case it IS Bristol Ridge, and not Summit Ridge, I think that makes it even more realistic to be soldered, as BR is a more expensive part.

??? How so? The 28nm process is cheaper considering its maturity and yields.

[Notional]

Just now, Coaxialgamer said:

well bristol ridge IS bigger , but not necessarily more expensive . Wafer costs have gone way up since 28nm

Either way, if you zoom in, it looks like 1 core, not 2. Only that the solder splits up for some reason.

[Notional]

1 minute ago, patrickjp93 said:

??? How so? The 28nm process is cheaper considering its maturity and yields.

Meant Summit Ridge (Ryzen) of course.

[Lewellyn]

Quote

16th December

EDIT: 
Out of curiosity I googled 7700k delid and saw a number of websites reporting on my findings of the 30 degree temp drop, but were unsure on cooling method in both tests so thought I should clarify: The before and after tests were 100% identical (with exception of delid) to isolate the delid, both tests used:
 

• The same cooler at the same setting: Kraken X62 - Fan @ 50% and pump at @ 65% (by the way, this cooler is only about 3 degrees cooler than the H110i, but has better config available)
• Exactly the same BIOS settings - which was a x50 multiplier and voltage offset with LLC level 2 resulting in a stable 1.344v under prime 95 load
• Memory stock (C15) and Uncore stock.
• Prime 95 v27.9 run for 15 minutes in both tests.
• The same case / config / install (tested with case side panel installed - i.e. how i would actually use it for 24/7 use so the test would be more meaningful to me):https://s30.postimg.org/gzluzd7w1/Case.jpg

 

[VerticalDiscussions]

Can i delid myself ^^?

[Mr.Meerkat]

@wrathoftheturkey Hopefully you realised that he said:

21 hours ago, Misanthrope said:

AMD into green numbers

which means into profit  

[AlTech]

18 minutes ago, Mr.Meerkat said:

@wrathoftheturkey Hopefully you realised that he said:

which means into profit  

or you could say AMD is in the Black (meaning making money) instead of in the red (how ironic) not making money.

[Kamina]

Hopefully AMD's new CPU will have Intel invest more into their R&D and thermal issues.

[ConiferousJelly]

22 hours ago, Misanthrope said:

This might be evil but...don't fix this before launch. I'd like to see the 7700k be too hot to compete with Ryzen chips. Maybe we can finally get AMD into green numbers that way.

Exactly why they would fix it. Intel should start getting scared and wearing and actually do their best, that's what we need. We don't need Intel to fail, but to be challenged.