Intel's i7-7700K Kaby Lake Overclocked and Reviewed at Tom's!

[Lays]

1 minute ago, Simon771 said:

 

 

 

 

Well i read that article from der8auer. I also remembered that he is selling some delidding tool for CPUs ... I knew I heard that name before.

Thanks for all the explanation ... I now understand that i7 6700k and other 4 core CPUs can't be soldered because CPU die is too small.

But what I noticed while reading that article ... liquid metal performs about the same as soldering. Is there any reason why Intel wouldn't use that? I know it's not going to work with sub zero temperatures, but that's the same with TIM they are using right now, isn't it?

I'm realy tempted to delid my i7 4790k and put liquid metal on CPU die, and normal TIM around the CPU die (to prevent liquid metal from touching other small thingies around there). But I just don't want to crack my 350$ worth CPU just yet lol

Then again, 15°C lower temps would dbe great.

 

I promise I won't bash Intel about soldering i7 6700K anymore, but I will still complain why are they not using liquid metal, unless you can also convince me that there is some legit reason for that also

Probably not using the liquid metal because I don't think it lasts forever, pretty sure it has a "golden window" of use, it'd probably need to be replaced in the long term, the TIM they use right now lasts a reaaaaaaal long time.

 

Probably also would cost loads more, adding cost onto us. 

[Simon771]

3 minutes ago, patrickjp93 said:

Because Intel isn't allowed to. Liquid metal TIMs contain mercury or gallium which Intel legally is not allowed to use anymore.

Oh wow, that's something new.

By doing 1 min research it seems like mercury isn't good to use because it can poison you or something.

2 minutes ago, Lays said:

Probably not using the liquid metal because I don't think it lasts forever, pretty sure it has a "golden window" of use, it'd probably need to be replaced in the long term, the TIM they use right now lasts a reaaaaaaal long time.

 

Probably also would cost loads more, adding cost onto us. 

Ah well I will just stop whining, and buy i7 6800k next time ...

 

Thanks for all comments and explanation guys, and sorry for posting crap before doing research.

[patrickjp93]

1 minute ago, Simon771 said:

Oh wow, that's something new.

By doing 1 min research it seems like mercury isn't good to use because it can poison you or something.

Ah well I will just stop whining, and buy i7 6800k next time ...

 

Thanks for all comments and explanation guys, and sorry for posting crap before doing research.

Same with Gallium. Both are toxic heavy metals.

[porina]

2 hours ago, MageTank said:

Stable isn't a word that deserves a prefix or suffix. You either are, or you are not. Yes, specific workloads cause instability more often than others, but it does not change the fact that in the end, you were unstable. That is the problem with the current world of overclocking. We have a bunch of people running around with pseudostable overclocks and try to pass them off as if they are "24/7 stable". The stability is always black and white, it's the users definition of stability that seems to differ per person.

I'd go one step further than that: no one can claim they are 100% stable. You know when you are unstable when you get an error, but an absence of an error for some length of time doesn't mean you are stable. It is impossible to prove.

 

As an example, my first Skylake build seemed fine initially, but I was running Prime95-like tasks at PrimeGrid. After some time I worked out I got occasional errors on bigger units that hit the ram (FFT sizes up to around 3M or so). After successive bios updates the error rate went down from an average of one or two a week, down to one every one or two months. Low, but not zero. You try diagnosing at this error rate! In the end I replaced the ram, and the problems went away. Someone not running something stressful 24/7 with error detection probably wouldn't have noticed those errors and declared the system 100% stable. Even now I can't say my system is 100% stable, as I haven't tested all possible workloads. I think I saw in some early reviews, the video coding engine was more sensitive to overclocks. I never use or tested that.

[MageTank]

18 minutes ago, porina said:

I'd go one step further than that: no one can claim they are 100% stable. You know when you are unstable when you get an error, but an absence of an error for some length of time doesn't mean you are stable. It is impossible to prove.

 

As an example, my first Skylake build seemed fine initially, but I was running Prime95-like tasks at PrimeGrid. After some time I worked out I got occasional errors on bigger units that hit the ram (FFT sizes up to around 3M or so). After successive bios updates the error rate went down from an average of one or two a week, down to one every one or two months. Low, but not zero. You try diagnosing at this error rate! In the end I replaced the ram, and the problems went away. Someone not running something stressful 24/7 with error detection probably wouldn't have noticed those errors and declared the system 100% stable. Even now I can't say my system is 100% stable, as I haven't tested all possible workloads. I think I saw in some early reviews, the video coding engine was more sensitive to overclocks. I never use or tested that.

Yeah, I try to tell people that stability isn't something we can quantify. As much as I have repeated it in this thread, i'll keep doing so until people believe me. Stable until unstable. That isn't just a random saying, it's an inevitable fact. Time impacts everything, and your overclock will eventually require more voltage to maintain down the road. It will eventually become unstable. Or you subject your overclock to a workload it was never tested for, and fails. Plenty of things that can potentially make a stable system unstable. 

 

100% stability is a myth. Something many have claimed to see, but have no proof of it. The funny thing is, this is just us talking CPU stability. Once you get into memory overclocking, stability becomes a cruel mistress. Leads me to my new motto: It's always the IMC unless proven otherwise, lol. Don't know how many countless hours I've spent fiddling with tertiary timings just to hope and pray I don't crash randomly a week from now because no test exists to accurately 100% show perfect ram stability. People that consider memtest86 a stress test in that regard disgust me, lol. 

[Citadelen]

18 hours ago, Space Reptile said:

-snip-

As opposed to Extreme Edition? That's great.  

[michaelocarroll007]

21 hours ago, zMeul said:

from an i5 to an i7? yeah I would

not really what that comment was aimed at.... Apples to Apples 

[MEC-777]

Will be sticking it out with Haswell for another 3-4 years, looks like.

[Castdeath97]

17 hours ago, MageTank said:

For mobile users, it's an entirely different story. The much higher clock speeds on the mobile SKU's offer additional performance at the exact same TDP. It's essentially a free overclock for the same amount of heat produced, meaning the cooling solutions do not change. Assuming they get the power consumption under control, and it consumes comparable power to it's Skylake counterparts, it's going to be a good thing. Especially once you consider how many of the mobile SKU's do not support overclocking in the slightest. Additional CPU performance, no matter how small, is always welcome in the mobile world. I am personally hoping to see some Kaby Lake GT4e SKU's in laptops. 

Any integrated graphics improvements? That would be good to see in mobile.

[ace_cheaply]

10 hours ago, MageTank said:

 

Almost reminds me of the people that claim to be running 24/7 5ghz without a delid. 

Not 5ghz, but I've been running 4.9 ghz on my wife's 4790k 24/7 for months with no delid on a 120mm aio @1.28v without a single crash.  Crashes on load at 5ghz though :P.  If I did delid or have a custom loop maybe I could hit 5.  It did once for 10 min of burn test but then crashed and hasn't lasted more than 10 seconds since then, and I'm afraid to go over 1.35V with my shitty cooler.  It already had spikes of 85 with sustained temps of 80.  

[Scitesh]

On 11/29/2016 at 9:43 PM, Pyrotechnika said:

I'm honestly not too impressed. 200MHz upgrade but way more heat and power? Anyone with Skylake shouldn't think about upgrading.

Anyone with haswell  should not upgrade

On 11/29/2016 at 9:50 PM, Spork829 said:

So much meh. I'm rooting for Zen to be good more and more these days.

At least AMD did some work instead of just changing the numbers on the box

 

On 11/29/2016 at 10:06 PM, RideToHellBestGame2013 said:

Capitalist plot to push for more people buying AIOs!

 

*BREAKING NEWS*

INTEL ENTERS AIO MARKET DUE TO KABYLAKE BEING FURNACES

212 evo FTW

 

On 11/29/2016 at 10:30 PM, Sampsy said:

I'm just sitting here with my 5Ghz 2500k wondering if intel is ever going to release a product worth me buying. My PC is now nearly 5 and a half years old. This is getting stupid. 

I am  sitting with my 4.3Ghz 3770k

[Scitesh]

11 hours ago, MageTank said:

I like how people seem to think that you can just fix binning by throwing a decent loop at it. "Any CPU can hit 5ghz if you cool it hard enough". Sure, in some cases, that might be true, but then it begs the question, is it 24/7 stable while doing so? Let's face it. Overclocking is very black and white. You are stable until you are unstable, that is the most simplistic truth of it. What might be stable today, may be unstable tomorrow. I know people with different overclock profiles for the different seasons year round, lol.

 

If proper cooling was all it took to make 5ghz stable on the vast majority of CPU's, more people would be doing so. We don't see that because the lottery is real. No loop will make a chip that is incapable of 5ghz, do 5ghz, at least not without the risk of severe degradation.

 

Almost reminds me of the people that claim to be running 24/7 5ghz without a delid. 

LN2  for everyone

[S w a t s o n]

The King is dead. Long live the King. (ZEN)

[Jovidah]

19 hours ago, patrickjp93 said:

EXcept you're wrong, but hey, it's not just Geekbench, now TH has shown 3 more cases.

Just out of curiosity...but why are you so convinced it has IPC improvements? I don't think even Intel is claiming that. In all their marketing they're really focusing on all the other minor aspects that have improved (mostly useful for mobile and low-end CPUs). 

 

For all the haters... so far I think Intel has been pretty 'honest' about how this is just a minor refresh. Of course it makes no sense to upgrade to this from Skylake. Why would it? It's just 1,5 years apart. Upgrading just 1 generation really never made sense in the last... 15-20 years. It hardly ever makes sense for video cards either.

Personally I don't mind the somewhat slower pace too much. At least it means my PC won't be completely obsolete in 3 years. 

[PocketNerd]

On 11/29/2016 at 9:00 AM, Sampsy said:

I'm just sitting here with my 5Ghz 2500k wondering if intel is ever going to release a product worth me buying. My PC is now nearly 5 and a half years old. This is getting stupid. 

How are you cooling that bad-boy?

[PocketNerd]

18 minutes ago, Sampsy said:

Just a Corsair H60. It's been sat at 4.8 / 5Ghz for pretty much the entire time I've had it. While it hasn't been on 24/7 it has been used heavily. 

Wow.. makes me wonder just how high I could OC my 6600k with my 212 Evo...

[DELTAprime]

So for reference what type of OC results do people get on a 6700k on air? The 7700k @ 4.8ghz sounds impressive except that my 6600k is only 200mhz behind @ 4.6ghz.

[MageTank]

19 minutes ago, DELTAprime said:

So for reference what type of OC results do people get on a 6700k on air? The 7700k @ 4.8ghz sounds impressive except that my 6600k is only 200mhz behind @ 4.6ghz.

"On Air" is a broad spectrum. Are we talking 212 Evo's or TC14PE's? Different coolers make for different results, not to mention the lottery is still very real. Over on OCN, with a sample size of 92, the average overclock was 4.68ghz. https://docs.google.com/spreadsheets/d/1wQwYMGsSnMpKxrEesNVUSoP7hGykFWw1ygJsxwx64e8/edit#gid=0

 

The highest OC I see on air in that results, is 4.8ghz on a Noctua D15, using a very tame stress test and JEDEC memory (meaning lower package temps). 4.7ghz seems to be the average for larger air coolers (Noctua D15, Phanteks TC14PE's, etc) while 4.5ghz seems to be the average for 212 Evo's according to that list. Again, the results submitted in this test use different stress tests, so there are many variables to consider. What might be stable at Cinebench or H264, might not be stable during Prime95 or Linpack. The people in this sample were not given strict rules to abide by, as far as testing goes, so they were able to submit their own preferred stability results. Chips that are stable under Prime95 at say, 4.5ghz, might be able to run 4.7ghz while gaming and never suffer a crash, it all depends on the applications you plan on using the CPU for. 

[NumLock21]

Tom's did a retest

http://www.tomshardware.com/news/intel-core-i7-7700k-kaby-lake-overclocking-update,33119.html

[MageTank]

5 minutes ago, NumLock21 said:

Tom's did a retest

http://www.tomshardware.com/news/intel-core-i7-7700k-kaby-lake-overclocking-update,33119.html

Looks like the end result is the same. All of the performance seems to be coming from the increased clock speeds, nothing more. Though, it's nice to see the temperatures fall back to tolerable levels. Once these hit retail, we will see a larger sample size to get a better representation of Kaby's average overclock speed. Only then will we be able to tell if it's actually worth the investment of current Skylake owners, but knowing the 0% improvement in IPC, I wouldn't exactly hold my breath on a lottery. 

[NumLock21]

1 minute ago, MageTank said:

Looks like the end result is the same. All of the performance seems to be coming from the increased clock speeds, nothing more. Though, it's nice to see the temperatures fall back to tolerable levels. Once these hit retail, we will see a larger sample size to get a better representation of Kaby's average overclock speed. Only then will we be able to tell if it's actually worth the investment of current Skylake owners, but knowing the 0% improvement in IPC, I wouldn't exactly hold my breath on a lottery. 

Kaby Lake isn't worth the investment for current skylake owners. They're more for those who are looking to upgrade their 5+ year old pc. 

[Duskie]

I'll stick with my Haswell Xeon @ 4.1Ghz

 

I have a spare Z97 motherboard and a 4690k, so i'm set for another 3 years!

 

..Probably more, unless I decide Zen is too attractive.

[Kinda Bottlenecked]

Eh... I'll see how the price is with the 7700k. A XPower shouldn't be a problem for either. Or maybe there would be a refresh of the XPower. Mmmhmm.

[Lays]

On 11/29/2016 at 9:00 AM, Sampsy said:

I'm just sitting here with my 5Ghz 2500k wondering if intel is ever going to release a product worth me buying. My PC is now nearly 5 and a half years old. This is getting stupid. 

Wanna see something neat? 

 

http://www.overclock.net/t/1578480/i5-2500k-4-5ghz-vs-6700k-4-5ghz-in-games#post_24549500

 

 

HI @MageTank Suhdude

[silberdrachi]

Im glad now days that i really dont need to worry too much about upgrading my CPU compared to my GPU. Id like to get some of the new features that intel added onto the core, but its not required really for what i need. Otherwise id have to go all in and get a new cpu/mobo/ram and that could run as expensive as upgrading my gpu that i plan to do next year sometime.