Clevo Overclocker's Lounge

Yeah, Boiling ssd's. But +91C on PCH chips ain't much better. Put on the TRIPOD mess as well and you can see why people call it Failienware!! Amazing the Dellienware ship still is over water and not already have sinked down with the sharks, fish and whales.

 

Most people are not enthusiasts and don't even see what is going on.

 

Hope they are smart enough to put aside money for paying the extended Premium warranty when they customize the laptop. Some put all their hard earned money on extra or more costly hardware and minimum for the longest possible support/warranty.

 

Yeah... PM961 drive is... not good. Not to me, not in my DM3. Sustained load kills the writes heavily. Though I've not had it overheat like that.

 

Have you found out why clevo skimped on the KM1 single-pipe bridge heatsink fin design?

 

That's ludicrous. And, I here I am thinking that the low-to-mid 70's on PCH temps on my 16L13 was inexcusable (bare die, so yes... absolutely... half-assed and inexcusable) but that's really insane. If you told me the people at Alienware didn't know any better, I'd have to say they are lying to shirk their responsibility to sell something that functions correctly 'cause that's just messed up.

ANYHOO... This fixed up my 16L13 PCH temps real good...

 

It's no bigger or heavier than the XPS M1730, M17xR2, M18xR1, M18xR2, Alienware 18 or P570WM that I used to travel extensively with. I've taken a few trips with the P870DM-G that I had before this one. It has never been a problem for me to travel with a monsterbook. I actually like it. The expressions on the faces of those that have never seen things like these bad-ass laptops are always priceless.

 

The P870DM-G is very portable, the DM3, oh boy thats a totally different story lol.

 

In what way to you find them different? Just the weight? I guess if you're not used to it that could feel like a problem. The P570WM was like hauling a couple of cinder blocks in a backpack, but it never bothered me as much as leaving at home and wishing I had brought it with me. My last trip I took the 16L13 thinking it would be easier. It was easier, but before the trip was over I regretted not taking the P870DM3 instead. It always gives me warm fuzzies having a behemoth laptop in a backpack while traveling. Everywhere I go feels like home, LOL.

 

The DM1 feels quite a bit lighter than the DM3. Specially with 1 PSU.
The DM3 and 780W PSU together, are ok to carry around to and from the parking lot to the Home/Office, but on days when i use the metro, its pretty much like asking for punishment. One can definitely get used to it, but i find it excessively heavy to carry around every day.

 

Yeah when I compared my DM3 to my DM1 the weight difference from the chassis alone was very noticeable, but that's not something that bothers me since I've carried my M18x and P570WM around as well lol.

Just was surprised about the noticeable difference.

 

Woah// The controller chip reaches 105 C for me, but it doesn't BSOD...

 

It is recommended to use whatever produces the best results and to not be afraid to try something new. Doing what is popular may not always be the best approach, but that varies by clique, too. Kryonaut is a great product, but it is nice to have at least one option that is equivalent.

 

Cool. I have been using only gc extreme and ICD once for a friend. Just reading some old posts about ICD makes it sound like a bad choice but when I tried it it was not bad at all.


Sent from my iPhone using Tapatalk

 

I do not agree with people that consider IC Diamond a bad choice. I've used it for many years with overall good results. If the heat sink fit is sloppy (a common issue for laptops, especially Clevo,) it works better than anything else. It is also very durable. Gelid GC Extreme works well also, but I have not found it to be as durable as IC Diamond or Kryonaut. But, Gelid GC Extreme is a WHOLE LOT better than most of the popular pastes like MX and NTH1 (which I have found to be worthless).

 

Wow I did not know Kryonaut was more durable than gelid extreme. Is it more viscous than gc extreme?

 

Hey @Mr. Fox , I am actually considering re-applying my paste job, but this time with Kryonaut. Are there any downsides or things I should keep in mind for a P775?

 

If yoo have time. Take a look on this mess.


First...


After...



@Georgel Maybe do a test application first. To see if the paste spread evenly and not to much. Then apply the final paste (if you are unfamiliar with applying thermal paste).

 

Did he apply LM like one would traditional paste???!@

 

It seems so If people see this YouTube video maybe they will be less scared for applying all too much LM?

 

Anyone else here not like making contact with the back of the CPU like that?

 

Nope mine is flat after i lapped the heat sink.

 

What I meant was in the picture. The way he's getting his greasy mitts on the back of the chip.

 

Oh lol i totally mis read. Neah i never touch them like that. If by chance i do touch those contact, i make sure i clean them with Isopropyl.

 

Maybe he hoped for sub-zero temp?

 

It should be fine with the gold plating and point like contacts but not best practice no.

I don't think they know what best practice is.

He is using as much as you do with thermal paste on top of the IHS, let alone on the die.

 

I'm facing the same issue but I don't notice any throttling when playing games. Playing witcher 3 at everything max, cpu temp is around 60-65C with some spikes to 70C and gpu temp is around 65-75 while "drive temperature 2" would reach 80C max. Not sure what's causing this but even at idle it hovers around 50C to & 60C hotter than the CPU lol.

 

Interesting. Didn't do anything special here. Just my normal everyday 4.7GHz daily driver. Maybe I'll fire up the AC and max it out and see how it looks.



UserBenchmarks: http://www.userbenchmark.com/UserRun/4460473
CPU: Intel Core i7-7700K - 112.9%
SSD: Samsung 960 Pro NVMe PCIe M.2 512GB - 305.7%
SSD: BPX 480GB - 254.8%
SSD: Samsung 840 Evo 1TB - 101.6%
SSD: Samsung 840 Evo 1TB - 101.2%
RAM: G.SKILL F4 DDR4 3200 C16 4x16GB - 104.9%
MBD: Clevo P870DMx-(G)

 

That goes to show you how RAM helps.

My run was with 5GHz and your CPU portion of the benchmark slightly succeeds mine at 4.7GHz

 

I wonder why it refuses to bench the GPU on my system. Does it do that on yours as well. I ran it several time and it gets to the graphics test, runs a bit and stops. Then the online results exclude it. I thought maybe SLI was messing it up, so I disabled that and it still wouldn't measure GPU performance.

 

I've been out of the loop for a while, what is currently the best Nvidia driver version?

Sent from my Nexus 6P using Tapatalk

 

For overclocked benching? For W7 378.50 works best for me. For W10 my GPUs like 378.66 the best. Most drivers newer than those are kind of sucky.

 

Nope, mine runs fine. SLI on no issue. Maybe see if anyone else with the Prema mod has fault getting it to bench the GPU?

Sent from my OnePlus 1 using a coconut

 

It's over 9000! Coming soon in 5 years, the successor to Intel Optane

 

Nice photoshop? xD

 

With raided CPU connected storage you can get those sequential numbers.

 

What themral paste do you recommend the most then?

I will return and exchange mine with one from another batch to see if this brings on any improvements

Also ordering some SSD cooling solutions.

 

Kryonaut is highly recommended. Most people have had good results with it. My cores are all within 3C of one another on a 7820HK at 4.5 ghz.
It's important to make sure you put the heatsink on evenly and in a criss cross motion (partial turns) on the screws so you get an even mount. This will help prevent dry-out and keep the cores closer in temps. Unless you're using a jokebook tripod Alienware @Papusan

Some people did not have good results with Kryonaut. @Phoenix
There have also been some stories of tubes with very dried out paste, but those were usually 1g packages on amazon. I purchased the 11.1g version from MemoryC on amazon and it worked well.

Phobya Nanogrease Extreme is also good, better than MX-4, but I found higher temps and higher core differetials (6-7C) using Nanogrease instead of Kryonaut. @Mr. Fox had better results with Nanogrease Extreme, but I had worse results with it, and I tried using it twice. So I'll just keep my nanogrease as a backup.

 

Taming the Beast - P870DM3 7700K Lapping - Before and After Results
OK, you saw it on the Tornado F5/EVOC 16L-G-1080 and the @Papusan example. Here's a how-to video.

Spoiler: Open to see images from video demo

Rendering this YouTube video with the CPU @ 4.5GHz saw a max core temp of 66°C

@Rage Set @Mr. Spock @j95 @Jon Webb @Johnksss@iBUYPOWER @KKthebeast @Falkentyne @Papusan @dspboys @Phoenix @Prema @tanzmeister @bloodhawk @D2 Ultima @TBoneSan @GTVEVO @ssj92 @Dr. AMK @leftsenseless @Huniken @OWNORDISOWN @ole!!! @ajc9988 @Mobius 1 @iunlock @Stress Tech @hmscott @Ted@HIDevolution @Donald@HIDevolution @Zoltan@HIDevolution @thattechgirl_viv @ThatOldGuy @Eurocom Support @Meaker@Sager @Prostar Computer @Larry@LPC-Digital @HTWingNut @jaybee83 @Robbo99999

 

Damn thats a lot of tests!!
I lapped mine yesterday as well - http://imgur.com/a/fiDxZ

 

Imagine lapping both sides of the ihs and the heatsink down to 2500 grit. It did very well (was literally putting together mirrors). Great video.

One thing on lapping. I advocate a turn every 4-10 strokes. If the pressure isn't even, you can build in a slant to the IHS. Turning it causes any unwitting pressure or a bit of dig in at the front edge to be even per side. I also do straight line. Temps look good!

Sent from my SM-G900P using Tapatalk

Edit: also, on the wet dry paper, if you add just a little dish soap, it breaks the water surface tension. Everything gets wet and great contact between the ihs and sandpaper.

 

Actually, it's here. @meaker is right. There is now a pcie3.0x16 raid card you can buy with 4x Samsung pro 960s. Think of those scores!

Sent from my SM-G900P using Tapatalk

 

Still looks false, 960 Pro doesn't go above 60 4k low queue depth read; 1300 is a *huge* jump from that. And raid has 4k overhead too

Sent from my OnePlus 1 using a coconut

 

http://www.highpoint-tech.com/USA_new/CS-product_nvme.htm
https://www.techpowerup.com/233586/...101-series-pci-express-3-0-x16-nvme-raid-ssds

I'll try to find the article showing the latency doesn't matter and raid speeds up Q Depth to a degree.

Sent from my SM-G900P using Tapatalk

 

Dedicated raid cards are much better yes, but each *single* SSD does not pass 60MB/s 4K QD1-4 on its own. The shown screenshot set that number to 1300+. That's literally nearly a 6x multiplication on a single quad-drive setup. Even using 8 drives, it's still going to be nearly 3x the bandwidth. With the info I have now, this is into the realm of magic for performance boosts

Sent from my OnePlus 1 using a coconut

 

I'm not arguing that this was photo shopped. More that we are seeing much faster speeds now than ever, including other than Optane.

Sent from my SM-G900P using Tapatalk

 

Well, I'm always skeptical. The more I learn about tech the more I realize companies lie outright. Until something is out and proven I don't tend to put much care into it

Sent from my OnePlus 1 using a coconut

 

It wasn't photoshopped. My friend and I were comparing our 850 Evo drives. (We were gonna trade my M.2 for his SATA).

Those are the results I saw, they are that fast because I have samsung rapid mode enabled. So my ram is used as cache for the SSD.

 

I forgot about that option. I use it on my P770ZM.

Sent from my SM-G900P using Tapatalk

 

I forgot about it as well. When I saw those results we were both confused until I went into samsung magician and noticed it was enabled lol