KabyLake initial testing/batch

[nickolp1974]

ya that's kinda what I was thinking, might just switch for a different one, his seems really bad. I have a L631 also but havent seen anyone with one yet.

 

My moneys on the 631 being the better chip, get that Apex to work!!

[Alan_Alberino]

L639F977 5000/4500 with 3600 mems R15 requires 1,216v... It's without delid, used some cheap Arctic Alumina to test it worked but hitting 90 degrees, delid with kryonaut coming soon.

[nickolp1974]

L639F977 5000/4500 with 3600 mems R15 requires 1,216v... It's without delid, used some cheap Arctic Alumina to test it worked but hitting 90 degrees, delid with kryonaut coming soon.

 

Looking very good, gotta be sub 1.2v after

[Alan_Alberino]

Looking very good, gotta be sub 1.2v after

 

Hope so, need to test a z270 and then delid and back to MOCF, hope I have time tomorrow

[speed.fastest]

Hey all, did you try what voltage to pass 32m 5g/5g 2c/2t? Mine before delid need 1.425v to pass it, after delid it dropped down very big, so far 1.36v working very good, did'nt try lower yet.

[flanker]

L641G639 Will see what happens (at Friday). But I would like to have better L639...

[M.Beier]

Guys, would you please let me know your VTT for given uncore/ram when testing? I am trying to figure if that is holding my uncore back, or its only a matter of vcore/temp combination...

 

Also, RAM frequency, how your VTT is at given frequencies...

 

I can get 4900 Uncore bench stable, but for torture tests, the crazy ones, 4800 is better.

5000 I simply cannot boot at.

 

3600 on RAM is a nightmare, 3466 only 2T, 3200 1T is highest.

Hope some of you got time to test impact of VTT

 

Cheers.

[speed.fastest]

VTT is default at half of VDIMM, VTT only affecting memory stability (B Die might be need specific VTT for Ultimate Stability, every kit is different!), for more uncore you need more vcore or colder temp.

[Strong Island]

well at least after 1 cpu, Tapakah's batch# could have kept me away from this horrible thing. 1 of mine is L640F774, 1 number after his L640F773

 

Tested last night on Apex. 5.0ghz r15 needs 1.31v and 5.1ghz wouldnt pass at all, would even lock up in OS at 5.1 at 1.35v. Un-delidded on custom loop.

Edited January 18, 2017 by Strong Island

[subaruwrc]

tested 5 pieces of L638F9xx with deepcool neptwin air cooler cca 26c ambient.

2 did 5ghz cb15 at 1.30v

2 did 4.9ghz cb15 at 1.33v

1 failed 4.9ghz at 1.36v

Edited January 18, 2017 by subaruwrc

[R3dRumG33k]

On Asus m8AH 1.312 volts 51/46 1.125 volts /3600 4 sticks 1.35 volts, 100% stable threw Blenders benchmark test run that lasts just over an hour with the lid still on, still waiting on my thermal Grizzly to delid, Also on my bios chip for my Z170OC, But im still tweaking and tuning I think im gunna end up somewhere around 5.3-5.4 after delid and more voltage (at least hoping)

[alex_007]

L643G218 ; vid 1.136

 

I have same Batch L643G218))

My score R15 at 5.1 is 1111 by 1,36v

[x-rated]

L639F826

5 GHz AVX2/ADX 5B y-cruncher / HEVC Handbrake encoding stable with ca. 1,34V (CLU under heatspreader & MX2 under AIO H100i GTX)

cache set in bios to 5 GHz, but CPU-Z shows only 4,8 GHz max.

with 4x 8GB @ 3 GHz on Maximus VIII Gene

Edited January 19, 2017 by x-rated

[I.nfraR.ed]

My L643G I wrote about earlier, but after delid and AIO cooling, Grizzly on both, CB15

5.0/4.5 1.189V (measured with DMM)

5.2/5.0 1.328V

 

VID is 1.104 according to OCF board.

 

On LN2 and Z170 OCF - what I tested, it might be able to get higher with same settings

6.0 1.6V -100

6.2 1.6V -120

 

Then below -130 the "fun" begins. It starts freezing under load, nothing seems to help.

Full pot same 6.0GHz profile I saved freezes when I start a bench. Pushing VCC PLL and CPU PLL to 1.8+ helps a bit, but still freezes on higher frequencies, 6.2 no go, while it is ok on higher temps. I will try to get MOCF to test.

6.0 cache booted, didn't test if CB will pass, I wanted to check cores first.

Edited January 20, 2017 by I.nfraR.ed

[speed.fastest]

My L643G I wrote about earlier, but after delid and AIO cooling, Grizzly on both, CB15

5.0/4.5 1.189V (measured with DMM)

5.2/5.0 1.328V

 

VID is 1.104 according to OCF board.

 

On LN2 and Z170 OCF - what I tested, it might be able to get higher with same settings

6.0 1.6V -100

6.2 1.6V -120

 

Then below -130 the "fun" begins. It starts freezing under load, nothing seems to help.

Full pot same 6.0GHz profile I saved freezes when I start a bench. Pushing VCC PLL and CPU PLL to 1.8+ helps a bit, but still freezes on higher frequencies, 6.2 no go, while it is ok on higher temps. I will try to get MOCF to test.

6.0 cache booted, didn't test if CB will pass, I wanted to check cores first.

 

Some CPU scale poorly under cold, but before that did you check CB R15 core temp load at 6-6.2GHz? Maybe hit positive and can you run fullpot? For me always start doing heavy benchmark on Skylake (i know its different) at -135 Celcius to reduces cracking TIM possibilities, cracking possibilities higher on -100 to -135 Celcius.

[I.nfraR.ed]

6.2 -120 1.6V a little bit over 0, then full pot it stays under 0 when it actually passes.

Full pot no problem booting and restarting, even saving bios settings, then doing stuff in windows, CB15 and CB11.5 freeze on the 1st set of squares.

[speed.fastest]

6.2 -120 1.6V a little bit over 0, then full pot it stays under 0 when it actually passes.

Full pot no problem booting and restarting, even saving bios settings, then doing stuff in windows, CB15 and CB11.5 freeze on the 1st set of squares.

 

Do you try max CPU Clock? If it good maybe your CPU is very hot, need higher thermal conductivity TIM to make it pass under load. Try change TIM again then heat up the pot to 70 Celcius, but dont start the system! (It may cause dead CPU). Then try again under cold. Heating the pot make it harder to crack the TIM under heavy CPU load in my experience. And dont forget start benching on -135 Celcius or Colder. I hope it fixed your problem.

Edited January 20, 2017 by speed.fastest

[I.nfraR.ed]

I always pre-heat the pot to about 50 before applying tim.

Then mount, cool down, start the system.

 

And I always start running benchmarks at say -90, then gradually decrease temp and check if it scales. Didn't go full pot at once.

 

Today after first attempt I tried heating to 70 as Allen suggested, then cool down again, start the system, but still the same.

The CPU was very hot with stock tim before delid. I could not do much more than 5.1G CB15 because of temps. After delidding the temps seem ok.

Maybe I just need to try on MOCF with CB Killer volts. Will try to get one soon.

 

PS: Or it just doesn't scale under cold, as someone said about L643 batch.

Edited January 20, 2017 by I.nfraR.ed

[Guest Bullant]

One I bought 2 days ago think it's L641 batch will check when home, not delid 50/45 will delid soon and try cold

 

Room temps prob 28c

 

Tested this CPU quickly last night on ln2 ,6.6ghz R15,think I was 1.79v....Ran 1m 6.9.I also bought second CPU same batch and it was junk on water,this better chip has 1.169 vid the other junk one same batch was 1.232 vid on ocfm.

 

The junk chip was failing instant on R15 1.35v 5g no delid,was enough for me to pack it back up and sell

Edited January 20, 2017 by Bullant

[speed.fastest]

PS: Or it just doesn't scale under cold, as someone said about L643 batch.

 

I agree with this, some batch is just not scaling under cold, but it seems you hitting thermal limit. Want solder indium?

[Guest Sinebrychof]

Core i7 7700K L641G639 delid, runs very hot. MOCF

 

Cinebench 4c8t:

 

5.0/4.8 1.2v

5.2/5.0 1.29v

5.3/5.1 1.36v

5.3/5.2 1.42v, 5.4 core needs a bit cooler to pass than room based water as cores are in the 60s.

 

L641G631

 

5.2/4.9 1.36v delid normal water.

[shar00750]

get my cpu , not delid yet with h110 : 5225mhz 1.32v 4/8 past cinebench with maximum

74 degrees

[Alex@ro]

6.2 -120 1.6V a little bit over 0, then full pot it stays under 0 when it actually passes.

Full pot no problem booting and restarting, even saving bios settings, then doing stuff in windows, CB15 and CB11.5 freeze on the 1st set of squares.

 

There's your problem, -120 1.6V you never should see over 0.In fact with 1.6V below -60 you should not see above 0. So you have been skylake'd

[rtsurfer]

There's your problem, -120 1.6V you never should see over 0.In fact with 1.6V below -60 you should not see above 0. So you have been skylake'd

Ehh.. Not that far off.

 

Depends on how hard he went on the Cold bug volts. Not at -120, but somewhere between -110 to -100C, with the ln2 volts & 1.6Vcore, you can definitely hit early positives.

 

-60C 1.6V, you definitely hit positives & this is without the Cold bug volts on. With those on, you'd be in 20s to 30s.

 

Kaby is very hot. I am not the best at mounting, but this has been a consistent experience for me & for some other people I have spoken to. I'll give it a 5C variation for someone who's exceptionally good at mounting.

 

Sent from my Nexus 5X using Tapatalk

[suzuki]

Core i7 7700K L641G639 delid, runs very hot. MOCF

 

Cinebench 4c8t:

 

5.0/4.8 1.2v

5.2/5.0 1.29v

5.3/5.1 1.36v

5.3/5.2 1.42v, 5.4 core needs a bit cooler to pass than room based water as cores are in the 60s.

 

L641G631

 

5.2/4.9 1.36v delid normal water.

 

 

 

Vid ?

I see only gems for water

 

 

Sent from my iPhone using Tapatalk