2018 15 and 17 Have Arrived!

15 has 6 core i7 8750H with 1070OC

http://www.dell.com/en-us/shop/dell...e-15/spd/alienware-15-laptop/dkcwcfl15caseso1

17 had 9 core i9 8950HK with 1080OC

http://www.dell.com/en-us/shop/dell...e-17/spd/alienware-17-laptop/dkcwcfl17caseso1

 

I would love to see someone buy it and benchmark it, because expecting dell, this is totally goin to be a nightmare

 

Well tripod again...

 

lmao they stuck an ever hotter chip in the same space and use the same tripod heatsink design. Good luck to new owners!

 

Less than impressed. £1000 more for a CPU upgrade. Not buying the thermal improvement given an almost identical design.

Think it's going to be the 7700HQ R3 for me and just keep replacing it until I get a decent one.

 

The tripod has never been an issue. If there are too thick thermal that unbalance the heatsink then you cant fix it even with 6 arms. Its that the arms are fairly long which gives them less pressure. If the thermal pads are are the correct size then there is no problem. But they seem to have made the arms much thicker this time. So I expect more pressure as well.

Also a problem is the very small chip die size which makes it hard to balance a heatsink on top of it to begin with. Hopefully the larger die helps out. The TDP remains the same by the way.

 

Totally agree. When the cpu is not overclocked and stucks at 45W TDP. The larger surface of the cpu die and new vapor chamber design should provides better thermal than before

 

I agree with you. it looks like much stronger... But only tests can answer to this question...

 

Ehh if I was still waiting to get one, at this point I'd just wait for a real refresh with new gpu.

 

Ehh the tripod design is the biggest reason why AW suffers from core imbalance issues. even if someone puts too much thermal paste, if you squeeze your heatsink against the DIE, the thermal paste that's too much, will just be squeezed out and will collect on top of cpu or gpu pcb. because the cpu has 3 arms instead off 4, at every corner one, the top side of the cpu effectively core #0 and #2 where most of the time hotter than the south side off the cpu (core #1 and #3) @Falkentyne @iunlock @alexnvidia can confirm it for you. And because the top side of the cpu is not being held down as much as the south side, it was recommended to repad with thinner pads so the Heatsink can have a more uniform pressure (cause the thinner pads allow it to lay down in a more straight position) have you ever seen desktop with problems like this ? exactly because in a desktop everything is being held down by a minimum of 8 screws. even the dell xps series have 8 screws, previous revisions of the AW MX series or other pre-dell revision have 8 screws.

As long as AW keeps delivering heatsinks with tripod designs, their laptops will fail.

 

You cannot push a heatsink more toward the CPU die if there are improper thickness thermal pads between the heatsink and the components surrounding the CPU this is why heatsinks with 4 screws around the CPU die still can have a core differential. This is also the reason why a lot of units (if not most, hard to say) do not have the core differential anymore. Hell I had a larger temperature imbalance with my previous Alienware than with my current for this exact reason. Yet it had 4 screws around the CPU. It is also not helping that the cpu dies from Intel are so incredibly small compared to the gpu dies.

Desktops have 4 screws by the way. Not 8....... and desktops do not have to deal with surrounding components that need to be cooled by the same heatsink.

Also and I need to keep saying this. The GPU side of the heatsink and the fan mounting points all have influence on the pressure its not just those 3 screws. @iunlock has confirmed this as well. People are so focused on 1 part that they tend to forget the bgiger picture and in my opinion thats what @alexnvidia does which is also a huge critic of the excellent research that iunlock has done.

The only downside is that the pressure is lower from the type of arms used. It's like a leafspring and the pressure come from one side horizontally instead from the top if a springloaded method was used. Industrial design 101. They probably have done it this way because the space was limited and you kinda see it everywhere now.

So yeah bottom line, when they are shipped with the correct pads there shoul dbe no issue and my unit is a testament to this. Attached is the prime95 results out of the box.

 

they changed the design of the 3 arm CPU cooler. The third arm is much stronger now. We will see the results soon

 

but that's my point !!!! the cpu does not have 4 screws !!!! it has only three. i have pictues of the top part of the VRM's not even deformed because of the pressure. number one is: if the pressure on the arm is not high enough it will not press down, number two is: if there are not enough arms the pressure will not be uniform. problem one can be solved with a heatspreader (copper shim) but 2 cannot be fixed ! and that's the problem. Also you're talking about vrm's but if the heatsink does contact due not having enough pressure, your vrm's cannot even start to work. so again, i think you're messing the gpu side for the cpu side.

on my cpu ( where is has 4 screws according to you) i have a picture of how good the 4 screw pressure is you're talking about. if the pressure is uniform, can you explain to me how come that my heatsink on the south side of the GPU sloped upwards creating a gap of a few mm at the side where there are 4 screws. only the gpu has 4 screws, cpu even in the new model still has 3 screws.

desktops also have 8 screws, or is the gpu being held down by gravity ?

if you google for a videocard with the heatsink removed, you can see that the heatsink has to cool everythink. Die, VRM/FINFET/MOSFET chips, every chip is being covered by thermal pads. even if one pad is thicker then the other, a few mm can be filled up by a heatspreader, or the thermal paste itself (depends offcoarse on the viscousity) you say that units do no have the imbalance anymore, but if you take a close look at Reddit or AW Arena, there are still people receiving units with core imbalance problems. not locally but all over the world. this indicates that dell/alienware does not have their **** togheter. that a heatsink still can have core imbalance issues due improper use of the thickness, it indicates that the person who worked with the pads did not have enough knowledge to work with them. cause you can easily buy a 1,2 or 3 mm thermal pad and squeeze it to the form of a 0.5 or whatever you want. BUT only if you have 4 screw mounts to screw down the heatsink for a uniform pressure (but only AW releases a revision with 7 points to screw down a heatsink. FYI a desktop gpu and a desktop cpu combined have a minimum of 8 screws around the die at the edges to screw down the heatsink. desktops even come with an IHS so you're not screwing the heatsink directly on top of the die (like how we do in our laptops) but agains the IHS, so that the IHS can spread the HEAT. that's why it's called in Intergraded Heat Spreader, so the heatsink can suck up more heat from the cpu that's being spread around. the same can be done with shims on our laptops. coppers shims where also used back in the day when there was no fancy IHS like there is now

 

but what heatsink is that, 1060, 1070 or ?? and have they changed it for all models. it does indeed look sturdier, but you cannot keep a uniform pressure, with only three screw mounts.

 

Is that the new heatsink for 17 or 15? Will it compatible with r4?

 

I think you didnt get my point that the previous 15inch model, The R2 had uneven core temps as well for some users including me. This was caused by too thick thermal pads, yet there where 4 screws instead of 3 surround the CPU. You cannot push the heatsink all the way down if something too thick is there in between and this was originally the case with the current heatsink in the 15R3/17R4. Once they replaced the too thick thermal pad on the VRM's on the CPU side the core temp differences disappeared for most users.

The biggest problem is the small CPU die, before this all people including Iunlock thought it was something else that caused it, even considering that there was a bubble between the coating and the sillicon itself. The biggest fix would be an IHS because that would fix most of the issues. Because it is not just pressure it is also the balance and it is hard to balance a large plate on a small bump in this case the CPU die. That slope is not because of the pressure I think, since its next to those highly bend-able arms. Isnt your heatsink just uneven to begin with? The pressure from the arms next to it have no way influence on that part that sticks out. I mean thats what I see incredibly often in gaming laptops. Uneven heatsinks.

 

i did get your point but if you google for the 15 r2 core imbalance problems you almost get nothing.... do you google 17 r4 you have guides, people making money off dells bad work (nothing against Iunlock), people advising how to get a new laptop from the technical support, but i can't find almost nothing about the 15 r2.. you'll get a few pages, but with the 17 r4 it's 3 google pages and youtube suggestions, with the 15 almost nothing. even if i set my vpn to us for more searches, it stays the same. with the 17 r4 it was an epidemic with the 15 r2 it was just some occasional people that had problems. i had an 17 R5 never had problems, except for a gpu that stopped working but that was the GPU itself, not the cooling.

it does not matter how small the DIE is, a copper shim will always solve that issue. as long as the shimis in a uniform contact with the heatsink, you can have a shim that's the same height as al of your thermal pads but it will not cool. the arms where bended almost 15% back, that's alot, and only think what it did was enlarge the gap between the heatsink and die. if you have a big enough copper shim, you can overcome that the DIE is too small.

Thermal pads are not flexible ! they stay in the position where they where pressed the last. so if you have a 5mm thermal pad, and push it down to 3mm, it will stay at 3mm, it will not take it's normal form and become 5mm again.

an IHS cannot be done, because the cpu is not socketed. Dell should just get their act togheter and make a revision with 8 screws. the picture you saw was from my second heatsink. i have 4 heatsinks all exhibit the same behaviour, but i fixed everything by using K5-PRO and a copper shim. icm with LM conductonaught

 

No it's not compatible

 

i can't comment much on the new R5 tripod design, the top cpu tension arm does look thicker and longer. however it is a known fact that most if not all current R4 are flawed, whether you @rinneh like it or not and this is especially evident for those who own 6820HK or 7820HK CPU that can be unlocked and overclocked. once you unlock and overclock, cpu core temperature difference can hit as high as 15-20C causing 2 of the cores to overheat, reaching 90-100C easily. it's not even up for debate. if the current design is all good, they would not have made any changes to R5 heatsink. there are very reputable people in this forum (excluding those who self advertise and self promote of course) and many more can attest to this defect. Dell had changed the heatsink revision and part number several times throughout the course of R4 life and yet non of the revised heatsink can fix this issue. yes i agree it wasn't as bad as before, and maybe for those who own weak HQ CPU the difference is tolerable, but we are talking about HK unlocked and overclocked.

 

Good news for the people that want/need it. The new Alienwares come with fancontrol built into the command center. Its part of their new controller which also controls the alienFX etc.

Well they could do a very thin IHS but it would cost too much. Doesnt matter if it is socketed or not. What a nice middle of the road solution would be is a space similar to what AMD has on their bare die GPU's, those raised sides that are on the same level help to balance out the heatsink similar to what some users including me with the very old AMD Athlon 1400mhz thunderbird CPU's.

I repasted various 17inch models with HK's they all did not had the core differential with new heatsinks. The 7820HK vs the 6820HK was a different story though, not sure if it was just my friends unit and the sillicone lottery but the 6820HK ran much cooler, about 8c on the same clockspeeds. The 7820HK did get hot about 88c. but dont forget those chips are better binned. The 7700HQ's are technically the same processors but locked and they are locked because the binning is less ideal for overclocking. The only unit I saw ever with a 10+ CPU core difference was from the very first batch.

And ofcoure they could improve n the heatsink. They could it make more consistent because some heatsinks rolling off the assembly line now sitll dont work well. What I do think though, if you have an older board and use a newer heatsink, it often seems to still nto work, as if they suddenly source the vrm's or something like that from another vendor with different tolerances.

 

the very thin IHS you're talking about is known as an copper shim, and costs a few cents on amazon, or ebay.

 

I think you are not getting my posts. I am not saying that the CURRENT mobile CPU's can be delidded because they obviously cant they have bare-dies. I am only saying that it would be beneficial if Intel started using heatspreaders on their current generation so that those small dies are less of a problem. The Intel CPU's even though more efficient get hotter than the previous generations because the chip die is so damn small compared to the older Core I7's.

But there is not reason why a BGA cpu cannot have a heatspreader. BGA is jsut a mounting method and nothing more, it does not say anything about performance or the CPU properties whatsoever. In fact the CPU's apart from binning are IDENTICAL but the substrate used to put those CPU's on is different (FPGA substrate vs BGA substrate).

I have no clue how you get to the conclusion why you think that I mean that the current CPU's have a heatspreader?

Those "raised" walls are there to balance the heatsink on and protect the capacitors but also the chip die. Intel uses a coating which protects the corners on the CPU die which are terribly fragile. AMD did/does not. During the AMD ahtlon days, users like me did our own custom shims to protect those corners because heavy coolers could crunch those corners when you would bump the PC case and the CPU heatsink would wiggle on top of that extremely fragile die. See photo for reference. https://i.ebayimg.com/thumbs/images/g/dfIAAOSwDTNarNY5/s-l225.jpg

I didnt come crawling into the PC world yesterday. I have been overclocking, PC modding for 20 years now and even have been doing overclocking comps a few years back. There is really nothing you can teach me about PC thermals and best practices regarding cooling, no offense (and yeah also part of the Dutch Tweaker scene, actually was one of the mods until a few years back).

Second regarding the heatsink cooling & iUnlocks findings. I know damn well what he did, I followed his work, communicated with him, saw the videos etc and he is part of the reason that Alienware quickly revised the heatsink with different thermal pads to begin with. But there have been a lot of people including me who do not publicize all findings for the community etc yet still do a lot of testing, test various hardware and are even in contact with the Alienware team itself (I have been assisting a lot of users on this forum in PM to get into contact with the same team to get their AW15R1 and R2 issues fixed regarding the burning DC in cables).

iunlock stated himself he considered the thermal performance as on of the best in its class. As long if you could get the core differences down and I concur with that. I never discredited his finding and I never went against that. Its alexnvidia and him that cannot see eye to eye with that and had some forum beef a few months ago.

 

and that's the point, i was talking about current cpu's, no point talking about older cpu's when this thread is about a laptop with 6820, and a 7820Hk wich are all BGA aka they never can have a IHS. Intel is not going to use a Heatspreader, else they still would be selling socketed cpu's. i already stated a few times that NOT- socketed cpu's have an COPPER SHIM as HEATSPREADER, there is no point for intel to add a shim in between the die and cpu, when the user can do it itself. and it would not only complicate the assembly of the laptops, but also slower, more expensive and prone to error's. sorry to say this, but you could have 40 years experience and still know jack ****. people interpret things the way they want, and do things the way they want. it could be that you are saying or using something wrong for all the years you have experience, nonetheless i respect everyone's findings and i like a heated argument. it shows that someone can stand behind their points of interest.

that Alienware has the best cooling in the industry is a fact, but that Alienware has one of the worst Q&A since dell took over is also a fact.

but the problems of the 15, where nowhere near widespread as the 17 r4 has. i'm not talking only about the heatsink !!! (notice how i said a few posts back already: problems)

the 15 r2 had core imbalance issues, DC cables and sometimes the battery's would not work correctly, the 17 R4 has, CPU or GPU or both, Screen, Battery, AlienFx, MoBo. those are the components that could come from the factory with problems, those are not even yet the problems described in depth.

also BGA is not JUST a mounting method. BGA has to do with science, and affect a cpu or gpu pretty dramatic, as a overclocker you should know that :
BGA packages are used to permanently mount devices such as microprocessors. A BGA can provide more interconnection pins than can be put on a dual in-line or flat package. The whole bottom surface of the device can be used, instead of just the perimeter. The leads are also on average shorter than with a perimeter-only type, leading to better performance at high speeds.

High density[edit]
The BGA is a solution to the problem of producing a miniature package for an integrated circuit with many hundreds of pins. Pin grid arrays and dual-in-line surface mount ( SOIC) packages were being produced with more and more pins, and with decreasing spacing between the pins, but this was causing difficulties for the soldering process. As package pins got closer together, the danger of accidentally bridging adjacent pins with solder grew.

Heat conduction[edit]
A further advantage of BGA packages over packages with discrete leads (i.e. packages with legs) is the lower thermal resistance between the package and the PCB. This allows heat generated by the integrated circuit inside the package to flow more easily to the PCB, preventing the chip from overheating.

Low-inductance leads[edit]
The shorter an electrical conductor, the lower its unwanted inductance, a property which causes unwanted distortion of signals in high-speed electronic circuits. BGAs, with their very short distance between the package and the PCB, have low lead inductances, giving them superior electrical performance to pinned devices.

if i want, i could write a bot for HWInfo forum that helps people PM the right people, but i'm not talking about that. i'm talking about having A solution (guiding someone towards the solution is not a solution). @iunlock @alexnvidia did some testing that involved measuring for example how much V or ampere there where running in to their cpu or gpu. it was @alexnvidia that came with a solution and with the results he collected, where as from you i haven't really seen a post with a solution for any AW problems.

also do i find it very interesting that you came up with a solution from AMD that they used in a architecture more than 10 years ago, when even ryzen does not use the walls. If you have so much experience, would be very interesting to know how those things are called then. every component, that's being used has a name, i could not found out the name of the "walls" you refer to. And i also find it very strange that AMD hasn't used them for almost a decade now.

i like this discussion, it put's things in another perspective.

 

You make a lot of assumptions.......

A soldered heatspreader similar to what was found for the last time in Sandy bridge and some socket 201X CPU's with flux was ideal for cooling, as it says it spreads the heat, gives a large footprint for thermal dissipation and it helps ironing out unbalanced heatsinks. As you well know current Ryzen CPU's still have soldered heatspreaders on which Intel is unfortunately cheapening out for years now even with their latest enthusiast CPU's and thats why people delid those CPU's for better thermals. That does not mean they could still implement a good heatspreader on the current BGA line. Did you check other laptops as well that they all have similar issues with the same CPU's? Pretty much any laptop except for the twice as thick MSI's have issues with the latest HK processors and it is all due to the small die size, imperfect contact of heatsinks etc out of the box.

A shim is not the same, thats just an extra layer which you connect again with thermal paste. The original heatspreaders where soldered on. It is very different. You do create a little bit more pressure but also stress points on the fan mounting points and heatpipes.

Ryzens have heatspreaders now yes, AMD has been using heatspreaders since the Ahtlon 64. But not on their GPU's they still have those spacer/shims whatever you want to call it. they have no specific name for it as far as I know also on some of their APU's which are used in consoles. Why do we even talk about this? Its just an idea I am thinking of and even of executing to make such a shim and test it out if it balances the heatsink. Its what worked back in the days on bare die AMD's. See attachment for details before you assume that I mean something else again.

Regarding FPGA and BGA in these cases intel does not make any use of specific BGA's advantages. Its just to make systems cheaper and thinner. But the chips itself are of the same design as their desktop coutnerparts and come from worse yields. This is well known. Just maximizing profits. ALso half your text is about chips with those "side legs" not vs FPGA. But actually they currently use LGA/Land grid array. Which does not have the limits of possible pins as FPGA. But instead of using solder balls you push them down onto grid contacts. So no there are no more interconnects.

Again why are we even discussing this?

The AW15R1 and R2 series had multiple issues yes, and I communicated with Alienware about the express charging issues and how bios 1.2.2 was pulling too much power at that time. Documented it and got in contact with them about it. Assisted other users as you can see in those topics about what to do and who to contact if out of warranty. There was not more than for other users next to say that they need to turn over express charging. I am not running a shop with custom cables but at least they know whats up with their laptop, why and what needs to be done. Also some AGA bugs where ironed out because of direct contact with Alienware. But its not a pissing contest who did the most for the community right?

As long if someone provides opinions with a foundation, then I will take that person serious.

 

15 R4 internals.

Does anyone know why in the UK Dell store you can only order a 60Hz panel?

 

There is a chance that the alienware 15 r4 and 17 r5 has solved most issues with the alienware 15 r3 and 17 r4. I think it is dell's thread to make a rubbish laptop design and then make a revision that fix the problems of the previous ones with the same design. A example would be the alienware 15 r1, alienware 17r2 and alienware 15 r2, alienware 17 r3. So Basically alienware 15 r3 and alienware 17 r4 is crap and hopefully the alienware 15 r4 and alienware 17 r5 is the fix. From my opinion tripod heatsink is not ideal at all. I don't understand how difficult is it for dell to make 8 screws heatsink. ITS MIND BLOWING. hahahahahaha... Still not worth to upgrade without the new gpu line..

 

I notice quite a lot of complaints from users that not used these machines first hand though and just repeat hearsay.

 

This isn't going to cut it and I expect we're about to see another repeat. The issue is where the pressure is being applied. In this case all of the pressure is being pushed down towards the chip and the arms flex because all the energy is on one side of the motherboard. This is in part why they use so few screws because the heatpipes get in the way.

The solution is to use the GPU approach used by AMD's Vega. Vapor chamber on top with screw holes and a curved backplate retention mechanism. That way the pressure of force is not all top down instead it would be coming from behind the BGA socket towards the transfer plate. Then repeat at each individual contact point either with another pressure mounted backplate or just straight screws since the thermal pads don't need a great deal of pressure to conform to their surfaces, only the BGA chips do.

This would make repairability harder because they would become more Mac-like with the retention screws behind the motherboard which would require removing it completely to repaste however it would provide rock solid contact.

Link listed for context:
http://www.tomshardware.com/reviews/amd-radeon-rx-vega-64,5173-3.html

 

errrr.. I have tried using the alienware 17 r2 and r3.. also the alienware 15 r3.. have not tried the 17 r4... Repasted the 3 machine myself.. the 15 r3 heatsink was not even due to uneven pressure because of tripod heatsink.. Sorted the 15 r3 as it was not a K cpu and because of balancing the thermal pads.. The dell thread revision thing that I was talking about was just me making assumptions.. ^_^

 

Are the gpus that are overclocked Flashed that way or is it software based overclocked. Interested in knowing how they differ.


Sent from my iPhone using Tapatalk

 

You don't have core differentials because you're barely pushing 60℃. When you get in the 80s is where the core temps start to separate.

 

Well, both ROG G703 and HP Omen X 17 have core differential issue. You can check their thermal results on Notebookcheck. They all have 4 screws for CPU heatsink.

 

Nope it doesnt because when i play cetain games and go full load it touches the 80c and the core temps still stay together. Again have posted various screenshots various times.

 

I had a quick word with them, and they said that they are expecting the 144hz panel to arrive later in the month. There is an availability issue and they are prioritizing the US market.

 

If i specifically ment you I would have quoted you

Bios I presume. Also on the current models they have been overclocked. CLean install on mine machine yet the 1070 @ stock boosts quite a bit higher than reference clocks.

 

Holy shmoly new gen is expensive...
It seems it is better to grab last gen at lowered price.
And again same HS design...history repeating...
Hmmm...this makes me think their engeneers are some family memebers of the Dell owners...

 

What kind of difference we talking. Got a last gen/current gen. I heard a statement that the Alienware 1070 is full fat desktop equivalent.
Was kinda hoping it was software based overclock. I just bought it last night. But £500-£750 difference told me save my money. And wait for Volta. Enjoy the current one for a few months and resell current one for prob same money I paid for it...or more.


Sent from my iPhone using Tapatalk

 

Wait for volta, if you dont need a laptop at this day.

My 1070 runs at about 1800mhz standard out of the box. While standard it should be between 1443 - 1645 on boost.But there is still room to overlock. The highest I got was 1949mhz and gave me these 3dmark results. https://www.3dmark.com/fs/15120340

No clue why the core speed is registered incorrectly in 3dmark though.

 

Imagine if it was it’s clocks achieving that. But it’s still good.
I have bought the aw15 r3 already. Got it used but perfect condition.
I don’t know if I should be disappointed I got a 60hz ips g sync or 120hz tn gsync would of be better. Coming from an 13” OLED. What would you choose?
Anyway having it now will help me know if I will be happy to have a 15” spending more on it. See if I can get used to it’s size, if I can’t. If there is no refresh for the 13/14” then I’ll consider the aorus x3/aero 14 Volta configs as they was nearly my choice instead of a aw13 if it wasn’t for OLED

 

I would go for the 4K 60hz, not even gsync because battery life is for me personally more important. But it is hard to go back from such an exquisite screen as found in the 13inch AW. Still no outlook on 15inch Oled panels as well unfortunately.

The 120hz panel is nice if you like competitive shooters but the contrast is not near that of Oled.

I myself steer clear of Gigabyte/Aorus products after all the **** I had to deal with when using their products.

 

yes but those are only two. MSI, Auros, Eurocom, Clevo and the brands that use a clevo based machine did not have any problems.

 

Just jumping in to confirm that my Eurocom machine, with a 8700k in it, even if it's delided, doesn't have any major core differential (from 1 to 3 degres under full load)

 

Most Clevo's in the same class (so not the super thick and heavy desktop CPU models). Aorus as well and I actually had those in my hands and could test various Aorus laptops myself. Most crappy brand out all of them.

https://www.notebookcheck.net/Aorus-X5-v7-i7-7820HK-GTX-1070-UHD-Laptop-Review.257858.0.html < 96c under full load in the stress test for aorus

https://www.notebookcheck.net/Asus-...-GTX-1080-Full-HD-Laptop-Review.287134.0.html < crazy tha tthis thick Asus laptop reaches 97c

https://www.notebookcheck.net/Schen...17-Clevo-PA71HS-G-Laptop-Review.284844.0.html < 17inch clevo running at 90c under load

https://www.notebookcheck.net/MSI-GS73VR-7RF-7700HQ-GTX-1060-4K-Laptop-Review.226124.0.html < THe 17inch MSI stealth with the 1060 touched 89c which is also not that great for the GPU in use and is in the same class as the AW17R4.

It's clearly a current issue for msot of the laptops out there, big and small brands.

 

why do you keep adding things from more than 10 years to the story ??? i was not talking about sandy bridge or any part that's older then 2 years, my laptop was made in Juli 2017 and i have an 7820HK and a 1080. you have a 7700HQ with a 1070. that setup did not even have problems, yet you're talking about a setup you do not even own...

first a heatspreader is a mostly COPPER made thermal component that sits between the DIE and Heatsink. an IHS is just a fancy name for a copper heatshield, where the cpu is embedded in. if you want, you can make your own Heatspreader. just solder a piece of copper against your heatsink to raise the Thermal Transfer Surface. you're talking about socketed cpu's when the cpu we have are BGA, and BGA cannot have an IHS because the cpu can not be embedded within the COPPER heatspreader. just read the difference; INTERGRATED HEAT SPREADER versus HEATSPEADER.

Copper extrudes heat, and the way our laptop heatsinks work are with forced air cooling. we where talking about cpu's and stuff you bring up gpu's !!!!! we where talking about CORE imbalance issues of the tripod heatsink. and even the picture you provided proves that the tripod let's the heatsink sit in an ANGLE. the fact is that there are companies that use adequate cooling, and some that use off the shelve parts, just to be able to sell fast and quick.

regarding BGA, i know intel is not using it for those reason, but that was not the reason why i quoted that part about BGA. i reacted on the fact that you said that BGA is just a mounting method, and you bring up something totally irrelevant about intel, that every AW owner knows that used pre-dell pc's or even comes to this forum. Myself replaced my cpu and gpu in my older AW laptops.

I never said it was a pissing contest, that's what you're thinking right know. I only believe what i see, and i haven't seen your name come around alot, or even being mentioned like Mr. Fox, Prema, Iunlock etc..
but i do see you comment some good stuff, and some stuff (about the tripod) that got me irritated. you do not even own a 7820HK and a 1080 to test the systems or even benchmark. everything below a 7820HK or a 1080,
is free of problems. you're locket ad what 3.4 ghz, while i'm running 1.1 ghz faster than you. thats a heck of a lot difference, and heat i need to dissipate in relation to your setup.

again, if you've handled a 7820HK and a 1080 stock machine from factory, i'd like to see your face when you see the results. three days ago i saw someone bought a 7820HK And a 1080, and reacted to umar on Reddit, because his system was overheating and thermal shutting down. it's not the pads or heatsink that's the problem, it's the 7 legs that are the problem.

 

you're reacting with laptops that or do not have the same price point or have the same specs i mean really ?? what do you expect from laptops that costs almost half of a AW 17R4 7820HK and a 1080 ???

 

yessss because those machines rock and go hard !!! i should'e bought an Eurocom, but in the netherlands a setup with a 7700K (At that thime) and a 1080 would cost me almost 3k, and i only payed 2k for my current laptop next one will most definitely be an eurocom !

 

I think you need to recheck the pricing because all are in the same ballpark or even higher. The Asus and Aorus are well in the 2000 and during launch nearing 3000, the MSI was at launch 1800 euro in the Netherlands and the XMG depending how you configure it. The Alienware depending on how you configure started from 1700 and the 17 around 1800. So how is that half?

 

the 17 r4 with a 7820HK and a 1080 started at 2500 with a 1080p screen that's the setup i have. those laptops you mentioned are almost all stock versions. the stock version of the 17 started at around 1500 euro's if you would cheapen the **** out of it.

a laptop with a 1060 will never cost more than 2000 euro's so go figure

 

Out of curiosity, during those reviews, did they limit the temp of the GPU to help the temps of the CPU?

I'm asking because earlier this year, I bought an AW 15R3 with the doomed 7820HK and a 1070 only to return it after having the laptop for 5 business days.

During those 5 days, the laptop was faulty out of the box, a tech came in 2 days later to replace the whole heatsink, we repasted and repaded with kryonaut yet, as soon as I would slightly OC (4.0 on all core) the computer would have the craziest core differential even while undervolting the GPU and CPU, so I simply got a refund.

I'm writing my story because I don't believe I have the same clevo base as the one in the test you linked. My Sky X4C is really huge and the thermals are really good

 

Ok final try to explain it to you.

The heatspreader's main selling point is that it is SOLDERED DIRECTLY TO THE CHIP DIE. That mean perfect heat transfer. Do you know why a heatspreader as done in the sandy bridge days is beneficial? It was soldered with Indium, Inidum is 8 to 10 times more effective in transfering heat than good thermal paste, 81.8 W/(mk) to be precise. Conductonaut comes close though but the gap is larger (73 W/mk)

Also explain to the rest of the world why a heatspreader cannot be implemented on a BGA chip? Because you know, there are examples where that has been clearly been the case, remember those good old PS3 RSX processors?

Why I brought up the heatspreaders? It is basic physics that it is easier to balance something on top of something big and wide instead of something small. Just as it would be easier to cool something wide with spread heat instead of heat concentrated on a very small point. Simply just putting a copper shim between the die and heatsink only makes the pressure higher but its nothing more than that. It does not spread the heat since you know you only make the heatsink thicker.

And again the BGA CPU's are exactly the same as the desktop CPU's in terms of the chip dies only the bottom substrate is different. BGA does not have anything to do with the thermal performance of these chips vs other solutions. Which you kinda tried to say you did? Dunno I have no clue why you brought up some basic wikipedia copy/paste about some basic BGA stuff which does not always apply.

I have repasted various recent AW models of which some contained the HK cpu's and they all ran fine except for the very first Skylake that I got my hands on from a friend and that one needed the full treatment of heatsink replacement. Again using 4 legs surrounding the CPU is not a recipe for succes of the other parts surrounding the CPU are higher and or if the thermal pads surround the CPU are too high. There is only less pressure but if you repadded and or have good thermal pads to begin with you DONT NEED the extra pressure and that was my point since day one. iUnlock stated multiple times that the cooling is best in class as long if you can balance the heatsink and he has proven it with his results. So yeah you like to throw around with his name, than make sure you also have read/heard his opinion. Just as that uneven heatsinks are already well documented issues in the gaming laptop world for years now with and without "tripods" (this is also confirmed by the review links i just posted).

Also again the 17R4 has the benefits a pair larger fans and a heatpipe extra compared to the 15inch versions so that would kinda make it all even if with the 7820HK version you know. Also you need to factor in that my fans barely run and only kick on during games but at a low RPM. So with a higher fan noise it still can be kept cool.