*** MSI 16L13 (Eurocom Tornado F5)/EVOC 16L-G-1080 15.6" Owner's Lounge ***

Man, thought I'd come back to say I miss this laptop. I've had several since I sold mine, including a high-end desktop with an 8700k OC'd to 5ghz, and I almost regret getting rid of it. I know it is tempting to want to upgrade to the latest and greatest, but you guys should hang on to this special beast! I picked up a GT73 6RE to ease the pain, so I'll be playing around with that for the next couple of months. Praying it doesn't have the dreaded 1070 GTX issue.

 

Ugh 6RE :/ That's the slug 6820HK CPU. 6700K's were hot running and slow enough (only golden ones could go past 4.8 ghz).
Not to thread crap, but i'm rather impressed with some of these MSI jokebooks. 4.9 ghz stable in PUBG at 1.375v (true voltage, since I set IA AC DC Loadline to 0.01 mOhms rather than the intel reference 2.10 mOhms, yes, this setting affects STATIC voltages for some reason on these jokebooks), and a 195W TDP GTX 1070 is actually impressive. I would actually be able to run at 5 ghz with enough voltage and game but the Zener diodes simply shut the laptop off at that current draw (why? because MSI coded these jokebooks to automatically power off, if the requested amp draw would guaranteed 100C core temps at maximum fan speed on normal thermal paste. Doesn't apply to the 16L13 though).

Fun fact: EC RAM register D7 set to an "01" or "81" also triggers the Zener diode trip. Higher values (00=80 btw) cause a 'countdown" like it's counting down to doomsday.....i kid you not....seriously.
(Setting D7 register to 01 on 16L13 doesn't 'do that).

 

Yup, waiting on it to arrive but I plan on applying some LM with new thermo pads and undervolt. If I can't keep the temps under control or it has the 1070 issue, off to ebay it goes. I got an extremely good deal on it and I know from experience there wasn't much difference between the 6 & 7-series i7's from an overall performance standpoint. Thermo.. that is an entirely different story.

 

I plan to keep mine for the foreseeable future. It's treated me very well and looks brand new. Not a single scuff or scratch that I can see.

 

Yes, the metal shield is there tied with screws!

 

Remove those four screws then disconnect the metal shield then test again. I think the metal shield can not be fully removed. Just remove the screws so it is no longer pressing against the heatsink!

HIDevolution removed those screws and the shield due to testing by Mr. Fox, showing that the connection from videocard to CPU made the CPU run too hot.

 

Ok i will test it! It connects well to the cpu gives heat from the gpu to it... on the other hamd as far as i remember it gives extra plus 10c to the gpu... however it would be still 80c..

 

I have removed the screws and put two small pieces of toothpicks under the metal shield The cpu temp has improved! The gpu 76-78C was (tested only for 20-30mins) the cpu @ 4,4Ghz war 82C max on 2 cores. On 4,5Ghz@1,19V almost was 88C... I will test more, however so far it seems that 4,4 can be managable. Let us see after some hours of games thx for the help, it can be also possible that the delid is **** under the ihs, but i need to wait until I have a spare 7700k to test...

 

What game?

 

cod ww2

 

Same, I plan on keeping mine for as long as possible for the foreseeable future. This thing is powerful enough that if I treat it well and maintain it, I can't see myself needing another laptop for many years to come. Maybe have to replace fans during that but if that's my only issue I'd say it's a pretty solid machine. I really should order some spare fans in case mine die since other people have had so many issues.

 

Lol my CPU fan died in 6 months

 

My died around 14 months, 2 months after the warranty ended.

 

Yeah mine will be coming up on a year old this summer so I feel like it's getting close to that time that I should have a couple spare fans just in case.

 

MSI replaced the fans on mine when they replaced the motherboard under warranty. These appear to be different to the originals.

 

I have two spare fans at the ready if/when one or both of mine fails.

 

Do they sound different? I can't see the photo.

 

Its a job to tell, my laptop has been away for a couple of months. MSI had to wait for a new board and factories were shut for Chinese new year.
It looks to me now like the PABD19735 fan comes in a couple of different flavours - the B which is plain plastic and BM which has a metal cover on the side you see when it is fitted. The PABD19735BM is the one that is used in the GT62VR retail unit. I now have the BM's rather than the original B's.

 

And the BM model is a direct, drop-in replacement for the old fans?

 

Please tell us the complete model of the fan

 

Man, you rock!
I have been testing the laptop, and removing the metal shield screws helped a lot! Finally the cpu stable at 4,4Ghz@1,159V, and the temps spikes max around 83C. These are rather spikes for few secs, otherwise the cores sit around mid70s. However the gpu temp has increased a little, tops around 76-78C. Now I am glad, that I have not flashed the 1.05V bios, I can vision that the gpu would be +10C... now my bios max at 0.9V - 125W. I will install battlefront 2, which is the true test for laptops, keeps the cpu+gpu @ constant 100%
I will give a feedback as well! What I am also considering, to buy a modded bottom cover. Could you share your general view about that? I have read it helps to the cpu (question by how many Celsius), on the other hand gives a hard time to the m.2 ssd and PCH...
thx in advance

 

Please give @Donald@HIDevolution and @Mr. Fox credit for this, not me.
They're the ones who found this out. I'm only a messenger.
They deserve all of the credit.

 

Maybe were getting closer and closer to me putting an i5 8600K in my Eurocom Tornado F5

https://cdn.discordapp.com/attachments/423032538207092747/425717058014281737/cinebench-5Ghz.png

 

You guys with the 16L13, please go into your bios and set IMON SLOPE to 25, and see if you can draw unlimited power from your 7700K CPU without reaching TDP (91W) power limit and have power limit throttling please (you would need to test a high overclock with something capable of pulling >100W, like FMA3/AVX prime 95.

 

visited today on Reddit:
https://www.reddit.com/r/intel/comments/85waz5/5_ghz_on_cinebench_8600k_z170_chipset/?sort=old
and


To start modding we need at least a bios copy and I don't see anywhere online,
or we must buy a programmer and make our bios backup (copy) and then reflash with modified me region and added coffelake's cpu microcode.
I'll wait (very interested to this thing) a bit more to buy ^_^ anything

I'll try later and hope to not explode!

 

VERIFIED!
Setting IMON SLOPE to 25 bypasses the 91W TDP restriction on the MSI 16L13 and lets you draw up to the EC shutoff trip of the CPU (Whatever that is!).

 

Testing..

It works. I just punched out a 1079 CB run at 4.9Ghz all cores. Zero throttle due to power, but temps jumped to 90s at end of test. Keep in mind my 7700K isn't delidded and I gave it a random voltage so will have to continue to tweak for stability.

 

Guess I'll have something to do tonight.

 

Please set core IA AC DC loadline to 1 (both values) and then re-test. (DO NOT under ANY circumstances go anywhere NEAR the upper limits, PERIOD. Auto=210 (2.10 mOhms) according to what I've found on some Linux coreboot posts (Z370 reference value is also 2.10 mOhms).

You may have to change the CPU STATIC voltage to compensate if you are not stable.
MSI boards seem to use VID BOOST which boosts the voltage signal on load--the more load, thej more voltage boost, on BOTH adaptive and static voltages if this value is left at the auto setting (reference value is 2.10 mOhms, which is a bios value of 210).

I just confirmed that eVGA boards function correctly and this setting is IGNORED when static voltages are used, which is how desktop boards perform. The MSI boards applying this to static voltages is irregular.

 

Upper limits of what? Upper limits of voltage?

 

No, upper limits of IA AC C Loadline.
Using a value of 62.49 mOhms (62.49) is a great way to send your motherboard into LGA heaven (or BGA hell).
Reference value is 210 (2.1 mOhms).

Try a value of 1. Go to the same place that had your IMON SLOPE, Core I/A Domain. Change IA AC DC Loadline to 1 and 1.

I tested 400 (4 mOhms) just for fun, and at 1.0v static voltage at just 4100 mhz, I was getting enough current and heat to estimate that the CPU was getting around 1.3v (!). Enough for FMA3 small FFT prime95 to shut off the laptop immediately (Because of EC limit hardcoded protection).

 

Tried it, got near-instant P95 failure at 4.7 GHz, even upping the voltage override. Dropped back to my old 4.5 GHz, rock solid. I shall not pursue this any further, so please do not ask.

 

@saturnotaku that's no problem. Although your results pretty much confirmed that MSI is using this setting improperly.
you don't have to mess with it anymore. But knowledge is free, and you should still know "what" is happening exactly.

MSI is boosting the voltage signal by "around" 120mv at full load. That is what the IA AC DC loadline setting does.
basically, it is a measure of resistance. 2.10 mOhms is the intel reference setting. The problem is, this setting is only supposed to apply to ADAPTIVE voltages, which acts like a "loadline calibration" type setting, to keep the laptops stable at full load, with a mild overclock.

Unfortunately, MSI, for some reason, decided to apply this setting to static voltages also, when it was not designed to be used with static voltages. @Vistar Shook already confirmed for me (he changed this setting from 0 (auto) to 1 (0.01 mOhms), that his eVGA laptop ignores this setting when on static voltages, which is standard behavior for desktop systems. So basically, on a MSI board, a static voltage of 1.3v, with a default IA AC DC loadline setting, will boost the voltage signal (depending on load), by as much as 150mv, so that a prime 95 small FFT (AVX disabled) load might give around 110mv of increased voltage, and a AVX enabled (or FMA3) signal (this draws maximum amps) could be as high as 150mv.

Some MSi boards may not be stable with an IA value of 1, 2 or 3 (this seemed to apply to @sirgeorge 's 6820Hk SLI GT73VR system)
If I may ask, what was your stable setting for vcore at 4.5 ghz?

@aaronne may be willing to do some tests for me since I already helped him with his power limits.

 

1.225v, IA loadline settings of 1 each.

 

Wait what exactly did you do to make prime crash?
I thought you were using a default (0) setting for IA AC DC loadline, and you changed it to 1 just now and then you were unstable.
That's why I made my last post.

So you basically tried to use the IMON slope=50 and tried 4.7 ghz and were unstable, ok I got it.

Aaronne was able to run cinebench at 5 ghz after he changed IMON slope to lower than 100 as he no longer had power limit throttling. (he gave me permission to post his SS).

 

It was unstable at 0 or 1. There are probably other settings I missed somewhere, but again, I'm not going to bother any further. I did tweak around with undervolting the GPU, and got some excellent results. At stock, the system would "fail" the 3DMark Time Spy stress test with a score of 94%. Lowering the voltage and boost clock to 950mV and 1900 MHz (max boost speed is 1911, apparently), respectively gave a boost of 3-5 fps during the stress test and improved its rating to 96.8%, which is just short of passing. Keeping the clock frequency at 1900 and lowering the voltage to 925mV allowed the GPU to remain stable, provided another 1-2 fps increase, and the system hit 97.6% in the stress test.

 

The ones I now have fitted are PABD19735BM-N395 0.65A 12V

 

I went a bit O.T.T. with my bottom cover holes at 76mm and the M.2 and PCH do get a bit warmer - due to there being less air sucked over them from the front vents. If I am using the laptop on the cooler things are fine but if I want to use the laptop away from the cooler, I have made some rigid plastic disks which fit snuggly into the mesh to block the holes (I never bothered putting any of the bottom cover screws back in so it is easy to whip off). Modded bottom cover is definitely recommend if you are using a cooler.

 

Thank you! As I have read 5cm holes are proposed... and anyone uses liquid metal directly on gpu or cpu-ihs? Is it worth doing? Definately I will make the bottom mod, but I am not sure with direct CLU... anyone experiences? thx
ps: I am not using any cooler... I do not like to have something big additional

btw: do you have any firestrike result? I am curious what is your laptop's score, as it is the same as mine

 

I have not fully tweaked yet, as new motherboard meant I have had to redo BIOS mod. I just ran FireStrike and got 15897 (G=19184 P=13659 C=7797)

Just checked my screenshot archive and I had a tweaked score of 16531

 

Thank you. Can you let me know the gpu's bios version? I am still confused the differences in voltage, watts, bioses for 1070.... my can do 18.000 points, and I have no clue why (and I have added +150mhz on core...)

 

Different GPU's for different form factors. All are 115W TDP for GTX 1070, but one them (not sure if its MSI) may be 125W TDP.
MSI 1070 "standard layout" version used in GT73VR (this is the version that can also be used in Clevo P870 laptops, with small mod to allow the stupid MSI tab shape thing to fit in) can use 86.04.5B.00.44 or 86.04.5B.00.3A and I don't know the difference between them, but TDP is the same. Can't be bothered flashing. Perhaps .44 version is used on 16L13? I don't know. Gt73VR comes with 3A.

86.04.5B.00.0D version probably works too, but I don't know. This was from GT72VR (different form factor layout).

 

Thx for the info... i cannot get how can be 1000 points difference...

 

Knowing my luck, the GPU in my laptop probably came from the polished turd department at MSI.

Is yours running at 1825Mhz base clock?

 

1825mhz boost clock
Plus 500mhz on mem

 

What is your base clock?

Sent from my LG-H815 using Tapatalk

 

1445 as far as i remember...

 

Anyone know anyone looking for one of these laptops ?
I have a special deal until end of the month, if anyone knows someone who is thinking about one of these.

I also have started recommending people take the warranty upgrades on these laptops because they might need to replace the fans, since we have experienced that at large being the main issue.

I wonder though, if anyone who uses a modified bottom panel has the same issues, my thinking is, the way the fans are restricted proper air intake is what causes them to burn out or die faster ultimately.

mind you have I have plenty of friends and customers who have this who have no issues yet and going on over a year or more. So it's not conclusive an issue that affects everyone.

Anyways crazy how this thread is almost 1,000 pages long hahah crazy.

 

So they were right. The fan itself is the exact same fan used in the Gt73VR that spins up to 5000 RPM. Only the fan housing is different.

 

dying to see F7 with possible CFL 8 cores