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

Yes, in my other response to the Creators Update question, I should have also mentioned that I probably won't see any "improvement" in Win10CU because only one of my apps has significant problems *now*. Since that app was released over 10 years ago, I doubt Microsoft gave it any special attention. (As a matter of fact, it actually does *work* right now, but has some cosmetic artifacts.)

 

I literally just got mine and am on vacation. All I've tried is a little bit of WoW, and one match of Overwatch. When I get a chance, I'll check my temps, and post back!

 

Thanks man. A good test is set overwath to custom frames and set slider to 300 and max out all the other details. If you leave it on the title screen for a bit mine will go up to about 86C and in-game peaks about 88C. Almost any other gsme that uses 98-99% gpu will make it peak at 90C.

 

Try tagging @Mr. Fox for help. He really knows this book inside and out.

 

I just delided my cpu yesteday and put in LM + kronaut. Also put LM on gpu and honestly, maybe I did something wrong or because it was my first time, I'm just pretty bad, I didn't see a huge difference in temps. CPU while gaming, say overwatch for 30min, I hover around 75C (-100mv, 4.5ghz). GPU is around your temps too, around the 90Cs. Fan maxed, no cooling pads tho.

 

Sorry for late reply, I use pangobright. it kinda give you the effect of dimming by greying out the screen. personally i think it works great

 

Yes, single gpu cooling on this laptop is a godsend once the cpu is dealt with (and I didn't even do much yet). That is why I tried SLI first. To specifically overheat it.

Sent from my OnePlus 1 using a coconut

 

Is a copper shim going to help? I don't think it will - if you add a copper shim you're creating an extra interface between 2 surfaces that has to be dealt once again with thermal paste - so you end up with 2 layers of thermal paste instead of one - that's gonna slow down heat conductivity. My understanding of shims - they're only worth using if your heatsink is not making contact with your chip, if it's making good contact you don't need a shim. If the heatsink is making good even contact with the chip across it's entire surface, and it's pasted up properly with good thermal paste then you should be getting good temperatures - I think you should return the laptop for a replacement if you're confident of the previous 2 points & you're still getting high temperatures.

EDIT: if you want to double check your GPU temperatures against a known standard, then compare your temperatures against the Tornado F5 reviewed by notebookcheck.net:
https://www.notebookcheck.net/Euroc...TX-1080-MSI-16L13-Laptop-Review.206710.0.html
Specifically, check your GPU temperatures on an extended run of Witcher 3, they show a pic of the GPU temperature under 1 hour of Witcher 3 stress. In fact, just Ctrl-F (find) "Witcher 3 stress after 60 minutes" on the page I linked above and you will see the pics of GPU temperatures I'm talking about.

 

My unit has 1070 instead of 1080 so there might be difference. But before undervolting, my temp would spike to 85c~90c. After undervolting and repasting, things sure cool down quiet a bit. For Overwatch, my temp is around 70c, doesn't matter how long I play. But I'll try your settings and post screenshots tonight.
Here's my temp like when playing dark soul 3. Room temp is around 70F/21C.

 

Please guide me to the secret art of undervolting the gpu!

 

Hi @Mr. Fox can you please take a look of my TPL settings see if those are right? I set P1 and P2 to 160000 in Bios, will stress test tonight see if that fix it.

 

oh i didn't undervolt GPU at all. Only did on cpu.

 

No problem, but how much are ur cpu undervolted to get minus 15-20c! The only time i got ur temps were when I unclocked my cpu to 3.7

 

If you place the @ symbol before a user's name it will notify them that you are talking to them. It's a great feature. When I use @Maxmoky the system should have sent an alert to you.

 

I get what you are saying, but I don't know if there is an issue. When I look at the heatsink I can't see much of a gap and the thermal paste looks evenly distributed across the top of the GPU core. I don't see how the shim could hurt and it would eliminate the theory of there of being a gap.

It appears some other users received their units so we should be able to determine fairly quickly if this is an issue with my unit or if they all run this hot under load. The issue with the Witcher 3 stress test you sent it doesn't tell the settings so I have no idea if they enabled hairworks. Anything under 98-99% gpu usage has the same temperature range. Once you max out the card that is when you see the 90C temps.

 

Somewhere on that notebookcheck.net site I think they outline the settings & passage they use for Witcher 3 stress testing (I did try briefly to find the details but couldn't find quickly - but sure I've seen something before). Shims will increase your GPU temperatures if your heatsink fits properly because you use thermal paste on both sides of the shim & therefore this gives you 1 additional layer of thermal paste - thermal paste is not as conductive as heatsink material + the interface between 2 surfaces is never perfect and therefore this slows down heat transfer too (with a shim you have one additional such interface).

 

Do you mind running Overwatch with max fans and sit on the title screen with max settings for about 5 min and then play a round against the AI or go to the training and just run around and tell me what yours tops off at temps wise? My CPU temps are perfectly fine, it is the GPU that I am concerned about. If you are hitting 90C as well that tells me it is the design of the notebook. Do you have Battlefield 1, Mass Effect, Nier or any of the other newer games that just came out? Firestrike does not run long enough even in a loop to get the CPU temps to spike. The cooling pad does absolutely nothing on this model, I have tried 3 (U2, Xpal3, and my custom made Antec one I modified).

Witcher 3 on Ultra settings without hairworks maxes out between 79-80C so still a tad higher than notebookcheck but they also had an i5.

 

Latest (and I think, permanent) FireStrike results:



This is with settings (see below) that I've been running for a couple of weeks with zero crashes. I think I'm about ready to call it "good to go" and stop fussing with things.

Relevant settings:

• All cores at 4.2 GHz non-turbo, 4.7 GHz turbo
• Core voltage static at 1.150 V
• Cache voltage static at 1.150 V
• GPU core clock at +150 MHz
• GPU memory clock at +300 MHz

During this run, my max CPU temp was 85C, max GPU temp 86C, fans in maxed mode.

I also measured power consumption with an external digital wattmeter, and the results were pretty interesting. All of the numbers below are total system power, not "package dissipation" from specific components like GPU or CPU. I am doing this because I'm interested in using the smaller 230 watt power supply for travel. I have one of these on hand, and so far the results have been good, but please note that the following tests were done with the stock 330 watt power supply.

• My machine idles at between 35 and 50 watts most of the time, with screen brightness at about 50%. The nice thing about this panel is that it's bright enough even in a well-lit office with LED lights directly overhead and an open window from the side (not glaring off the screen). 50% brightness on this 4K panel is very comparable to my external monitor.
• During the physics part of FireStrike, which heavily stresses CPU but not so much GPU, I saw loads in the 105 to 140 watts.
• During the graphics part of FireStrike, stressing GPU mostly, the power consumption of the machine varied around 210 to 225 watts.
• During the combined part of FireStrike, the power consumption was typically about 235 to 245 watts, with one brief excursion to 249 watts.
• Simply maxing the fans, without actually loading the system, did not measurably increase the system's overall power. We all know those fans consume some extra power, but the delta was not measurable against the normal up-and-down bobble as different background tasks start and stop. In short, fan power is not a significant factor compared to other components.
• The 330W power supply with no computer attached consumes fractional wattage (the power meter bobbles between 0 and 1).
• During my Unity lightmap bake operations (a CPU-intensive batch job that runs for a few minutes to several hours), power stays almost always in the 100 to 150 watt range (this job's CPU load is "mostly" maxed, but there are intervals where it uses fewer than 4 cores for a short time).

Finally, I did a more realistic load test with Fallout 4, graphics settings pretty well "ultra" across the board except that I backed off on god rays and depth-of-field due to personal preference for subtle rather than overt visuals there. Running in god mode and deliberately stirring up battles in an outdoor cell, near downtown Boston (i.e., lots of nearby objects, lots of NPCs, and lots of visual FX), I observed:

• Playing at 1080p, the system consumed around 135 to 200 watts.
• Playing at 4K, the system consumed around 215 to 230 watts.

As far as I can tell, the batter did not discharge at all during any of the above tests.

My conclusions about power so far:

• The 230 watt supply is just fine for day-to-day productivity with this machine.
• The 230 watt supply is not adequate for intensive gaming or similar heavy GPU load (no surprise there).
• The need for the 330 watt power supply is driven by GPU load, not CPU load (again, no surprise).

I plan to repeat a subset of these tests while running on the 230 watt supply, to see what will happen, but I wanted to post this much before doing that, in case I manage to cause a system shutdown that would lose all my text from the web form.

Disclaimer: Neither the manufacturer, nor I, will tell you to try running this computer on anything other than the stock 330 watt power supply. If you choose to do so, you are responsible for the results.

 

@syscrusher run msi afterburner and turn on the overlay to display gpu temp and memory/core usage. Run a game and verify it is at 98-99% system load and let it run with max fans for at least 10 min. Benchmarks is not going to show you max gpu temps, even looped.

 

If anyone wants info on how to unlock the bios purchased from HID pm me.

 

Point fully granted; my results are representative of my maximum real workload, not what the system could push if I deliberately do so. My lightmap bakes are a pretty good test (of CPU, not GPU). The last time I ran one of those, there was a period where the CPU was pegged at fully 97% or more on all four cores for about half an hour. My max CPU package temp then was 89C. (I was monitoring with HWinfo64, which tracks the peak.)

 

I re-ran some of my earlier tests, this time with the 230 watt power supply connected through the wattmeter. Here are my new observations:

• The power supply definitely does not impose a hard limit at 230 watts. I drew up to 245 watts from it for short intervals. Disclaimer: This was a bench test only; I do not recommend exceeding manufacturer specs. If you do so, it's at your own risk.
• FireStrike ran very comparably to the results with the stock power supply. For this test and for the previous one, I had FS run its demo first so it was a longer-duration load.
• Playing Fallout 4 at 4K, I noticed very occasional cosmetic blips on the screen. I hypothesize, but cannot prove, that the power supply may have a slightly irregular voltage when it's pushed beyond its spec. (Another reason not to try this at home.)
• The battery, which happened to be at 100% when I began these tests, was still at 100% when I finished them. I conclude that overloading the power supply, at least to the level I did so, results in the PSU producing more than rated power and not the battery discharging.

 

That's a very helpful comparison

How hot does the 230w vs 330w PSU at full load - 245w sustained?

Of course the 230w is gonna run hot, but it might run within spec even under maximum load.

I think I'd still prefer 330w, but if the laptop came with a 230w it would be nice to know it would work as a backup PSU just in case.

 

I don't own an IR thermometer, but I would characterize it as "noticeably warm, but not hot". I didn't run for a "sustained" period, only about 20 minutes. (In my world, "sustained" means one of my lightmap bakes at 3+ hours.)

Emphatically not. The spec on the PSU's nameplate is 230 watts output, period. If you do what I did, you're exceeding spec (assuming 95% efficiency). Do that, if you choose to do so, with eyes wide open and knowing something bad (even a fire) might happen.

I have two of the 330 watt PSUs, because I routinely work from two fixed locations. I'm using the 230 watt for travel, specifically when I don't plan to do any high-load work during the trip. Some of my work pushes this machine to the max, but when I'm just editing source code or administering systems from a bash shell, the system is running at anywhere between 35 and 85 watts, well within safe limits for the smaller PSU.

Some day when I'm doing VR development in a mobile situation, I'll probably have to lug along the 330 watt PSU again. I hope that by then someone will offer a smaller and lighter version of the 330.

 

We are saying the same thing, use 230w for a backup - and travel - and 330w otherwise to have enough headroom to run cool and safe long term.

The rating printed on the PSU is the use rating, but the maximum output is usually computed with a 10-15% safety margin, so a 230w should run safely up to 250w-265w, but I wouldn't do that as a sustained load.

 

@rancid

I did as you asked, but MSI AB wasn't overlaying, I'll have to mess with it later. Anyway, Overwatch, with in game FPS/GPU Temp/etc...plus GPU-Z and Core Temp running.

CPU temps: 80/82/78/82C after removing frame limiter and set to Ultra at 1080p with the 1080 GPU, 120hz screen
GPU stuff: GPU Core MAX - 1860mhz; Mem Clock - 1251mhz; GPU Temp MAX - 85; GPU Load 99-100%
According to Overwatch stats GPU Temp was around 82-84 (most of the time), about 160 FPS.

This was after roughly 10-15mins on main screen, and then a full vs match. Fans button at the top of the keyboard is ON!

 

After 5min at title screen ( max setting, max fan ), my gpu temps around 83C.
After playing a good 10-15 min, my gpu temp jumps to 90C =/... And after 20min? 93C. I think GPU is really running too hot, will probably do a repaste. Also, I also have bf1 and mass effect but those games is going to fry my gpu as of now, so I'll post you new results once I repaste. I somehow can't take screen shots in game sorry. =/ If im missing any info, just ask

 

This is actually extremely odd to me; OW's main menu (when the primary focus; not alt tabbed) should be far far hotter than in-game.

 

Um..i guess im ur opposite? I get super hot when ingame. Even after 5min in game play, I already hit 90C..im pretty sure I did something wrong when I applied some LM on the gpu

 

If your FPS is uncapped on the main menu (or did Blizzard do something today I wonder?) then the main menu is significantly more demanding than in-game. It should drive your temps to infinity and beyond.

Edit: maybe you're being TDP limited on the main menu and in-game you're not so you clock a lot higher and get more fps and thus your card gets hotter xD

 

Aaaah theres the problem. I capped my fps to 130FPS because more than that is just waste for me. Ill try to uncapped the FPs and see if ill get anything better/worst

 

I have a silly question: According to the manual, some of the keyboard keys with the FN shift are programmable to launch an application of my choice. So...how do I set that up? I didn't see anything in SCM or KLM. I'm fairly new to Windows, so if this is a "standard Windows thing" please don't laugh too hard.

 

Have you tried Shortcut Manager? I think it comes with F5.

 

Mine's from HIDevolution. I thought I installed all the utils from the DVD, but maybe I missed one. I'll check; thanks for the lead. (My stock Windows install ate itself, and I had to reinstall.)

 

I don't have gtx1080 but 1070 instead so it might not be useful. I took screenshots through my phone since overwatch doesn't capture on screen display. Played about half hour. I don't have max fans on. It's on Auto.

 

Adding a shim to gpu was a bust, temps were worse with a 0.5mm and a 1.0mm. Reapplied the thermal pads over the individual chips tonight and have to get some more paste tomorrow to put it back together. Can't explain the 5C temp varience this unit seems to have in some tests but not others.

 

Your laptop seems to be behaving like mine as far as temps. From HID my temp were anywhere from 91-94C. After a repaste yesterday they stayed mostly in the 89-90C range.

 

Realized I never responded to this.. sorry..

Yeah, it's glass on the surface, but interestingly the panel is semi-matte underneath (an anti-glare). It definitely had slightly better brightness and color (IGZO vs TN) but now that I've had this 4K for a few days I've gotten used to it and I've forgotten about the blade. This is a pretty good 4K panel IMO, I like it.

I just realized I had G-Sync on universally. I just ran it again with G-Sync disabled for the 3DMark executables and I got 16869 (that's literally 4x the Firestrike I ran on my blade in 2015), out of the box with zero tweaking from HID.

Obviously I leave G-Sync on with my games, the image stability is pretty amazing. It's truly a thing to behold at 4K i'm loving it.

 

@syscrusher Could you be so kind to take a screenshot of your XTU settings or did you migrate to Throttlestop?

 

What overlay software is that? RTSS+something?

 

ya the msi afterburner with rtss and hwinfo64.

 

Good morning, in my part of the world, to everyone!

Wanted to get some suggestions from everyone. I have about $100 that I wound up saving after the order was placed and some discounts happened!

I knew that I would be needing a new bag to carry this thing in and I have been kind of looking at a nicer mouse. Any suggestion on bags/mouse?

 

I'm looking for a bag myself, but I just got a new wireless mouse. I've had it for 2 months or so now, and absolutely love it. If you can find a sale it's around 115. Not on sale 150. Logitech g900. It only has 24-32 hour battery life, but it's usable while wired too.

 

@Phoenix and @syscrusher have some backpack recommendations for this machine.

 

So knock on wood, I think I figured out the temp issue. I decided to try a different thermal paste and that got my temps down to the 86C range under load. I noticed that several games were spiking the CPU up to 90C (I have a delidded processor) so I undervolted by -0.110V @ stock clocks and now my temps are 83-84C under full load. For those having issue, more than likely your CPU temps are playing havoc with your GPU as I don't think the thermal paste change is what did it (I used Gelid Extreme for those that care).

 

This is the one I got:



It is very durable, the handles are sturdy despite the heavy weight of the laptop with the charger, it doesn't feel like the handles are going to break if you hold it from only one of them. I have walked in light rain as well while having the backpack on and the laptop wasn't wet so that's another plus point.

Speaking of storage, you can store your entire room in there (not literally) but you have so many pockets / sections it's super roomy and convenient.

PS: It costs an arm and a leg but hey, it's the best out there and this beast deserves nothing but the best. HIDevolution has it I think maybe check with Donald

 

Sorry I'm always "pimping" you out for forum help... hahaha. Sorry, not sorry?

 

Do what now?

 

The Eurocom drivers can be downloaded from here: http://www.eurocom.com/ec/drivers(384)ec

Shortcut Manager is in the "hotkey" package.

 

hey guys, good news for those who wanted to run 'hackintosh', nvidia has released beta drivers for gtx 10 series. personally i bought the laptop as a replacement for both my dell and macbookpro and realize after buying that driver wasnt out so really happy now i can say goodbye to old mac!.