*** Official Sager NP9877 / Clevo P870TM-G Owner's Lounge! - Phoenix 4 ***

The smaller but stronger Delta fan won't fit. You can mod the house opening.

 

You guys are just exceptional in every way. And, awesomeness is so rare these days. It seems strange to me that it is so rare, because following the Golden Rule causes it to flow naturally.

 

I'm going to have to totally agree!

I only just joined the HIDevolution experience and I have to say "I'm totally blown away with how they are treating me there." And I mean that from the bottom of my heart!

 

I just hope above all else that the only changes they make is getting better. So sad to see good companies fall into the thing they tried to not be, I wish HID never does this. But I think as long as Donald has a say in it, they'll be good

Sent from my OnePlus 1 using a coconut

 

So I shouldn't worry then as long as one isn't overheating?

 

I wish a lot of consumers wern't so shady. There are a bunch of them roaming around trying to beat the system and if they can't...They start with the defamation of character riot act. Knowing damn well it was their fault from the start.

It's a two way street unfortunately..

 

Yes, for certain. I can't say consumers have no blame. It really is a two-way street.

Edit: @wtjwillis I came across a video here



seems like native 4K actually runs games better than DSR because of the forced filter DSR has...

 

I see he didn't want to activate the MSAA Instant Nvidia GPU killer.

 

MSAA is so bad on these deferred rendering engines, it makes me cry.

You think that's bad I can max out a 1080 by playing Runescape just by turning on MSAA. 160W guzzled for that old game xD

 

With AMD. It use to be a turbo boost in performance. Literally 2x the performance!

 

Haha sounds like how ARK treated me before.

Min spec: 34fps
Raise textures to high: 42fps

me:

Spoiler

 

can make metalic plates to direct air flow. even if its higher amp/voltage it'll likely because blades are small can only move so much air. the 5.1*5.3 is probably a better choice what do you think.

 

Yes that's right.

 

What if someone uses similarly threaded longer screws with a few washers? Or shorter ones?

They are an exceptional bunch.
When @Donald@HIDevolution said that HIDE was a wonderful company this is what came to my mind -

Wonderful is an understatement. When i bought my DM1, it took me a solid 5 mins of talking to Ted to know that YES, HIDE was the ONE i was buying my system from. This was after talking with other re-sellers and vendors for almost over a month! Thanks for being awesome!

 

Snip

 

As stated before... I would test both. What's printed in specs doesn't always show up in real life

 

That is still not going to fix the heat sink plate holes that sit lower than the actual heat sink, remember? That is the space needed to get rid of along with allowing it to fit with the bottom cover on.....but then I'm sure you knew this and you were just testing me.

 

i checked my measurement again, if we go for 3*3*0.5 can fit possibly two small fan side by side LOL, or one of 5*5*0.5. those two you found i donno where you even start looking, can't find anything at all, at least not 0.5CM thickness.

 

Ah yeah, neah it was me. Totally spaced out thinking about the heights.

 

John and anyone else who played with both KM1 and TM1, especially if you have the higher DDR4 Modules installed (3200MHz+):

1. What is the fastest stable clock speed for heavy benchmark loads were you able to crank from the DDR4-3200MHz, 1080SLI& I7-8700K? How much gain did you see compared to 2800MHz modules for the RAM? If you dont mind sharing the voltage you used and multiplier setting that would be great!

2. What is the maximum temperatures for both CPU/GPU/System do you reach when the laptop is under heavy benchmarking loads with the maximum performance settings?

3. How much better were you able to do in various 3DMark/Heavy games tests compared to KM1 system?

I appreciate the tips!

 

You do realize they benchmark using A/C vents on the cooling system etc, right? If you're trying to figure out regular overclocked usage, you're not going to get people running heavy overclocks for 24/7 loads usually. I'm saying this because you're saying "heavy game tests", which implies gaming at benching clocks to me, which simply isn't going to happen.

 

Definitely. I understand that maximum possible clock speeds for short sprints are tuned for short durations to just last the test. I was referring to maximum "stable" clocks for heavy usage (gaming).

Thanks for clarification.

 

I would say if you tweak it and do a lot of cooling work on a DM3, are in a fairly cool room (sub-25c) and get a nice low voltage for your CPU speed, I would consider 4.7-4.8GHz on the CPU and maybe 2000MHz on the GPUs (custom vBIOS probably needed) might be doable, but at full gaming stress you're looking at around 90c on the CPU as max spiked temps (probably mid 80s topped out) and low 80c ranges on the GPUs. That's my personal experience with this. But a TM1 seems to have a better cooling system overall and may be capable of gaming at 5GHz while the GPUs do their own thing, since they are not linked by heatsink anymore.

This is just me speculating though, and includes assumption of a lot of work and possibly liquid metal paste for cooling. I'm assuming constant over 90c temps to be overheating territory for this statement

Sent from my OnePlus 1 using a coconut

 

i only got 2400mhz ram so far, but cpu at 4.9ghz gaming dota2 900p with 1070 however i got a binned chip, voltage at around 1.2 to 1.22v underload, peaked 80/81c on auto fan profile which is extremely quiet. if you're gaming with faster ram and gtx 1080 x2 with much higher resolution i'd not go over 1.2v for cpu.

 

So I should be seeing better temps with TM1 with similar CPU/GPU clock?

moreover, would getting Raid 0 optane memory 32GB improve my 960 Pro small file transfer and loading speed?

 

My P870TM-G should be here in the next couple days, this is gonna be my first beast laptop
.

 

that honestly depends on your cpu. if you get an average one or a good one then yes no real change in temp, perhaps u might even some some improvement. i recommend HID because you'll likely need a delid and a decent chip to fully enjoy it imho.

i donno how raid 0 optane memory will work with 960, maybe better off to just get a single 960 pro and wait for 980/980 pro coming later. theres already a review out for 981 which means 980 should be coming soon, you can read the review at tomshardware.

 

I've been studying the MXM modules of the 1080, would anyone be able to post high resolution images of both the front and back of the cards? There SHOULD be a way to shunt mod the current sense circuit that determines the core TPL reported to the power management IC. You should be able to either pencil mod them, or just run a bead of solder over them to eliminate the resistance and fool the card into thinking it's only pulling a couple watts.

 

Pencil mod won't work
@Prema

if I'm correct, the resistance drop required to get close to 0 needs something with enough conductivity to drop it down The reason pencils work on "Volt mods" is because the resistance in mOhms is already high enough (like 6000 in something measurement), so you only need a SMALL drop to raise the vcore, example: 1.5v to 1.6v with a few strokes of a pencil. If you tried using conductonaut, you would destroy the chip because just one bridge swipe of LM would send the voltage skyrocketing to about 2.3v.

On the shunts, you're talking bout like .05 or something similar, and are trying to drop it closer to 0 (0 will put the card into "fault" mode and lock the core to 139 mhz), so a pencil isn't good enough. Neither is a conductive pen, because when the resistance is already that close to 0, you need moar powa to make it lower

 

I think the shunt mod on 1080 MXM has already been done and it worked. But, it did not make a huge improvement. It was better than stock, but not as good as modded vBIOS. If you search the forum, you can probably find where it was posted quite some time ago. And, it does nothing to solve the room temperature core clock throttling, which is the root of all evil with Pascal. Mods that let you overclock higher only to lose ground to that throttle cancer baloney are almost futile. That nonsense is what needs to be modded away the most. I really wish the NVIDIOTS had made the voltage and power behavior on Pascal the same as Maxwell. Other than MUCH higher core and memory clock speeds on Pascal, I think Maxwell was a much better product.

 

If I remember correct, it was @Meaker@Sager

Edit. Here + Here

 

I'm sure the shunt mod has been done, but it'll only matter gaming-wise at high resolutions. You could max the cards at lower resolutions and your clockspeed will still remain high, but higher resolutions simply hit the cards a bit harder in terms of power draw. This isn't a case of "well high refresh lower res vs high res lower refresh" either. The higher resolution rendering simply needs more power, even if at both 1080p and say 4k (however you arrive at 4k, be it DSR or internal resolution scaling or having an actual screen) the cards get pegged at 99% util each.

But the custom vBIOS would help the most. The shunt mod works so well for 1070Ns because they are anemic on the power limits; a 1070 has a 150W limit by default. A 1070N has more cores and a 115W limit on average by default (the highest being Alienware's 125W models, but some units as low as 100W), so shunt modding GREATLY helps its performance. But a 1080 has 180W limit by default (and uses up to 1.093v at stock), and the 1080Ns in these units have 190W TDP limit and only use up to 1.0635v maximum. So we have enough power for average use plus will use less voltage, and then you can always squeeze out some more with the MSI AB/EVGA Precision X voltage curve tweaking.

 

1070N Doesn't need a shut mod to shine. It just needs the right vbios mod. And a vbios mod is completely reversible while a shut mod has the potential to eat the solder off the board

 

Yes, very true. vBIOS mod is always going to be preferable. And I have heard of Liquid Ultra destroying the solder after a few months too. If one does need a shunt mod I very highly suggest Thermal Grizzly Conductonaut instead

 

They all contain gallium, which is what eats the solder over time.

 

Fair enough

 

What would be best is if the customer-hating imbeciles that design everything to be crippled, locked-down and flash-resistant would all suddenly die, get fired, or just go away or something, and be replaced with engineers that are hopelessly addicted overclocking enthusiasts rather than Nazi control freaks. Then we would not have to waste time on stupid work-around mods to fix their retarded messes. NVIDIA certainly does not run their business according to the Golden Rule, but with nasty bed-fellows like Micro$loth leading the way on customer-screwing, they're in good company... birds of a feather.

 

No don't give them overclocking enthusiasts, they'll clock everything slow so we can OC by 200%

Give them people like Donald who understand that it's a user's hardware and they should be able to do what they want. Like custom vBIOS mods. To fix Nvidia's messes. Like Maxwell, where I had to tell people to overvolt their GPUs when they played certain demanding games like Witcher 3 and streamed it at the same time, and they kept getting driver crashes, even at stock. And some idiots were all "you need to take responsibility if you tell someone to overvolt their GPU and it burns out". Like bruh if +50mV breaks a stock card it was going to die in 6 months anyway, plus how else are you gonna fix the crashes? I don't see you giving any plausible fixes, only me.

 

Is it true that Maxwell needs more voltage? Because I have a 980M with bad ASIC quality of 59% and I can overvolt it to 37.5mV and it felt lot stable than Stock Dell BIOS. Maybe I didn't notice any issues with stability thanks to Prema BIOS.

 

http://forum.notebookreview.com/threads/on-the-subject-of-maxwell-gpus-efficiency.773102/

and yes to both

 

Thanks. I think Prema has fixed most issues. Hell MEI is still unpatched on my system.
Do you know if MEI FW implements throttling mechanism and other performance improvements done by OEM? If all vendors use stock MEI FW from Prema MEI tool to flash it on AW on which MEI FW will be updated next month.

 

Unfortunately, this I don't know, sorry.

 

It can. It should not have undesirable throttling with the default ME firmware from Intel, but OEMs can use it to manipulate behavior in ways that enthusiasts do not appreciate. Intel can also use it to produce undesirable behavior, but I haven't seen an example yet (that I can recall) where they have.

 

I think @Vasudev have DELL/DELLIENWARE in mind. Download firmware/software from Dell’s servers is high risk sport.

 

Yes, you are correct. And, Dell/Alienware would be one of the OEMs to worry about as well. They have their silly laptop "skin temps" spec that is placed above everything else in the order of importance, including performance.

 

1: 3200 Mhz
Voltage is 1.35V and Multiplier is 133/12



2: Somewhere around the 80's to 90's if i just run fans on auto.

3: I never had a KM1

 

That has to do with DPTF driver.

 

Mr. Fox and Papusan: Some very very good( or bad) news. Windows 10 RS4 has more oil pastel looks and looks like Windows mobile. I'm trying the Insider preview in a VM and performance improvements to UI is good and Win Aero oil Pastel which is optimised to run in lower end CPU/GPU is consistent across most UWP in-house apps. Clean install Windows 10 takes 9GB as opposed to earlier editions of W10 which took 8GB.
No telemetry, fewer UWP apps and Win32 apps works great for now unless they decide to screw the public release version.
Oil pastels brings back memories.