• Guest, The rules for the P & N subforum have been updated to prohibit "ad hominem" or personal attacks against other posters. See the full details in the post "Politics and News Rules & Guidelines."

Confirmed - i9 9900k will have soldered IHS, no more toothpaste TIM

Page 33 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

xisor

Member
Apr 5, 2010
28
13
81
So no actual real world examples of a 215W power draw in non AVX loads then? Gotcha. Thought so.
anandtech got 220w on their asrock board, they retested with a different board and got a different result. There is a problem with board VRM throttling leading to inaccurate power and temperature results from some outlets, such as oc3d. gamersnexus and hardware unboxed are both digging into it further with extended power and thermal/vrm testing in the coming days. The primary issue here is most outlets got their 9900k's about 2 - 5 days ago. hardware unboxed has had his for several weeks now though, standby for further results as more testing is done. I do believe i've proven quite definitively that i was not infact simply pulling information out of my rear end, unless you think i'm clairvoyant, HUB's cover photo should clearly cover that part of it. Anyway, i will update you as further information is found sir. It could very well have been MCE results due to board vendors being naughty with reporting whether mce is enabled or not, gigabyte is guilty of doing that, see gamersnexus MCE debacle with the 8700k for that one. Need more data to confirm that. Either way the temperatures are indeed very problematic as i said, and der8auer has given us the reasons why. This is a very bad launch, with very bad pricing, and i don't see how anyone can be happy with what's happened here with this 9th gen product stack.
 
Last edited:
  • Like
Reactions: lightmanek

happy medium

Lifer
Jun 8, 2003
14,387
469
126
Im seeing about 120 to 150 watts at stock boost clocks in real world Non AVX applications in the reviews ive read..

Overclocking is a different story.
 
  • Like
Reactions: ozzy702

epsilon84

Senior member
Aug 29, 2010
996
704
136
anandtech got 220w on their asrock board, they retested with a different board and got a different result. There is a problem with board VRM throttling leading to inaccurate power and temperature results from some outlets, such as oc3d. gamersnexus and hardware unboxed are both digging into it further with extended power and thermal/vrm testing in the coming days. The primary issue here is most outlets got their 9900k's about 2 days ago. hardware unboxed has had his for several weeks now though, standby for further results as more testing is done. I do believe i've proven quite definitively that i was not infact simply pulling information out of my rear end, unless you think i'm clairvoyant, HUB's cover photo should clearly cover that part of it. Anyway, i will update you as further information is found sir.
Clearly the Z370 board that Anandtech had was erroneously feeding the 9900K with excessive voltage, they clearly said so themselves, which is why they retracted and redid the power figures in the first place.

Looking forward to seeing proof of this 215W non AVX load you're talking about. I'm not holding my breath though ;)
 

epsilon84

Senior member
Aug 29, 2010
996
704
136
Im seeing about 120 to 150 watts at stock boost clocks in real world Non AVX applications in the reviews ive read..

Overclocking is a different story.
Which is where I would expect it to be. Overclocking drastically increases power consumption of course, a 9900K is already pushed beyond its efficiency limits with a stock 4.7GHz ACT.
 

Zucker2k

Golden Member
Feb 15, 2006
1,185
531
136
9900k delid and test with TIM and liquid metal:
TIM

Liquid metal

Article: https://www.pcbuildersclub.com/en/2018/10/der8auer-delidding-the-intel-core-i9-9900k-results-in-much-better-temperatures/

The temps that Gamers Nexus got were in favour of the solder - up to 5c lower. The temps that der8auer found were in favour of the delid, both with liquid metal AND tim. Which is weird.
Der8auer is in the cpu delid business so I'll be a little skeptical of his findings. Intel hasn't suddenly lost the science of soldering.
 

french toast

Senior member
Feb 22, 2017
973
780
106
Clearly the Z370 board that Anandtech had was erroneously feeding the 9900K with excessive voltage, they clearly said so themselves, which is why they retracted and redid the power figures in the first place.

Looking forward to seeing proof of this 215W non AVX load you're talking about. I'm not holding my breath though ;)
They were not the only tester to reach that conclusion though..hardware unboxed? Also showed that.
If reviewers were reporting that that proves him right, at least initially.
 

epsilon84

Senior member
Aug 29, 2010
996
704
136
They were not the only tester to reach that conclusion though..hardware unboxed? Also showed that.
If reviewers were reporting that that proves him right, at least initially.
Please show me where HWUB/Techspot found a 215W non AVX workload. I've seen that review, and every power consumption graph is of the system, not CPU alone. For example:
https://static.techspot.com/articles-info/1730/bench/Power-1.png

The 9900K is ~45W ahead in system power consumption, so if that equates to a 215W load then the 2700X is a 170W chip also?! I don't think so...
 

french toast

Senior member
Feb 22, 2017
973
780
106
Please show me where HWUB/Techspot found a 215W non AVX workload. I've seen that review, and every power consumption graph is of the system, not CPU alone. For example:
https://static.techspot.com/articles-info/1730/bench/Power-1.png

The 9900K is ~45W ahead in system power consumption, so if that equates to a 215W load then the 2700X is a 170W chip also?! I don't think so...
Must have been an AVX workload then, anyway just one review proves his initial leak right...even if it was faulty in the end.
 

xisor

Member
Apr 5, 2010
28
13
81
Must have been an AVX workload then, anyway just one review proves his initial leak right...even if it was faulty in the end.
its likely one of my original sources (not hardware unboxed) had mce on by mistake, just like linus had the power limiter on in his review thats on floatplane right now, so he reports 130w max load in a stress test full system, and 58c max temp on his aio. Mistakes happen, lanoc made the same mistake, and have since corrected their review to reflect accurate power/temp data. the offiical PL2 is exactly what i reported though, i was not making things up :).

Please show me where HWUB/Techspot found a 215W non AVX workload. I've seen that review, and every power consumption graph is of the system, not CPU alone. For example:
https://static.techspot.com/articles-info/1730/bench/Power-1.png

The 9900K is ~45W ahead in system power consumption, so if that equates to a 215W load then the 2700X is a 170W chip also?! I don't think so...
the 2700X system components draw about 30w more power than the 9900k system components do, which you can see by looking at idle power. That accounts for that discrepency sir.
 

Brahmzy

Senior member
Jul 27, 2004
583
27
91
LOL I just ordered a 9900K from Newegg for regular price today, tax free, free shipping. No bundles or price gouges. They were available all day.
 

LTC8K6

Lifer
Mar 10, 2004
28,523
1,569
126
LOL I just ordered a 9900K from Newegg for regular price today, tax free, free shipping. No bundles or price gouges. They were available all day.
You mean the "regular price" of $579.00 or what?
Because $579.00 is not really the regular price.
 

ASK THE COMMUNITY