Alienware 17 R5 (Heatsink)

Can't you in theory post the pictures of the OTHER heatsinks minus that 1?

Sent from my Moto Z (2) using Tapatalk

 

Please go into every link:


17r4
https://www.ebay.com/itm/264504807919
15r3
https://www.ebay.com/itm/254394529929
15r4
https://www.ebay.com/itm/254382828340
Area-51m
https://www.ebay.com/itm/264502806801

Regards!

 

Bought one for my 17r4. Thanks!

 

Has anyone that's gotten one of these been able to do some before/after testing? I'm still interested, I just can't bring myself to try one without knowing if it'll make a difference since the heat pipes seem to be the same.

 

I ordered one 2 days ago, he said all modification are done (I hope so), waiting to be arrived, then I'll do some testing and share the results.

 

I'm also waiting for some results/ logs.

thanks.

 

Any word on improved heatsink for alienware m15? I'm considering between buying an m15 with a 2080 Max q or a mag 15 with 2070 Max q and the temps might make a huge difference there... Would they take a custom order?

Sent from my Moto Z (2) using Tapatalk

 

Just received my HS Cicichen sent me the one with the modification. This have no heatpipe bent.

I Should be able to install this tomorrow so i will post result before and after by the end of the week.

 

Ok guys finally here the result before (oem heatsink) and after (Cicichen full copper heatsink)

Both test are made with Grizzly conductonaut and Artic Thermal pads.

See the picture i used Realbench cuz its the best stress test to test CPU and GPU togheter it smore effective then any gaming session.

So with oem heatsink CPU @ 4.4 Ghz, -0.135mv = after 1 hours 96c max temp with ambient temp of 20c. GPU = 65c degres and PCH = 91C degres.

With cicichen Custom heatsink 4.4 Ghz, -0.135mv = after 1 hours 84c max temp with ambient temp of 20c degres. GPU = 65c degres and PCH = 73c degres.

So for me its the improvement i was looking for around 10c degres and there we have 12c degres on the cpu.

The only cons are the thermal pads layout. Maybe i could get better temp with perfect fit but i did my best.

I hope its gonna help some of you!

 

i have to say also no more fan rush when working on the pc.... no more high peak temp!

So my next test will be 5.0 Ghz with external 1080 ti with my alienware amplifier.

I'm sure we should be under 75c degres on peak temp.

 

This is a good results for @4.40, but we need to see it @5.0, waiting for your next update.
did you used the thermal pads included with the heatsink, or your own? please explain more about your attempts with thermal pads, it effect the results for sure.

 

No i used artic thermal pads 1mm and 1.5mm. you wont need 0.5mm. But you will need 3mm pads for MOFOSET. Its Cicichen didnt fix the big gap but still better then the first HS release.

We have to work on a perfect thermal pads layout cuz i'm sure with the perfect thickness we can get better result.

But you know result talk by it self!

 

Thank you for sharing some results with the new heatsink.

Can you explain why exactly? I was curious and did the same test as you. 1h realbench v2.56 stresstest with 44x and uv -135mV.

In my test realbench generated 202w (CPU 95w, GPU 107w) in total. FarCry 5 generated 231w (CPU 65,56w/ GPU 165,44w) in total.






I don't understand why realbench now more effective for you? Realbench creates an unrealstic load on CPU. Which game use 100% usage over the whole time? And GPU only 107w? Sorry, for me is that not really effective, because games creates the most realistic load on your system than any benchmark. In a normal real-world usage you don't have often a task with 100% CPU usage over an hour. Maybe video-rendering, but that's not the purpose of the gaming-notebook. We want the best possible gaming-performance.

Honestly not really, because your test goes over an hour and unfortunately you don't show the avg values but we need them to evaluate and compare your results. I give you an example what i mean.



Your CPU drops in the stresstest on 800 Mhz, but how often, what was the avg clock in the 60 minutes? Same for the CPU package power. If your CPU-clocks drop, your CPU works less and generate less power (= less performance in this case). Logically you have with less power better temperatures. Maybe your temperatures are better, but we can't compare it.

In your 2nd log you have on your GPU in max 124.96w and max temp 64°C. I had in my testing nearly the same 121w but with 51°C in max. My min value was 39,23w, your's 7,40w.



Under 10w is just an idle-load. Not sure how long you record your idle temps and watts, but you can easily tweak your temps with it. I'm sure that wasn't your intention. That's why it's important where you start and end the HWinfo log. Can you do again the test? 15 minutes is enough.

For this test is that impossible, because already 4.4 GHz reached nearly the powerlimit of 110w.

 

The point is to test CPU and GPU togheter...( Cuz you know in a laptop cpu and gpu are on the same heatsink) Realbench reproduce heavy work load and gaming session. But you know you are so smart and better then evryone here... buy the HS and then you can make your own test and you will be able to share those test with ous. Peace man!

 

Who said i was going to try 5 Ghz @ 110w???? I'm not a geek but not a noob if i put 5Ghz in my bios i will put the power limit i will need! More power = more heat.

 

Is your Monitor GSync?
Could you please share your UV on IGPU/Unslice/CPU Cache and System Agent?

 

I've spent enough time and money in different mods, heatsinks and testing ( One of them). So it's normal that I'm a bit skeptical with aftermarket heatsinks, especially with this huge copper heatsink. Everyone thinks copper ist best solution and I thought it for long time too. In my testing was a copper heatink slightly worse than an aluminum heatsink. Thermal conductivity (and also heat-transfer) with copper is of course better but for passiv cooling i had better results with aluminum. Tested in different ways and cases with R4 and R5. That's the experience which I've made. I'd wish Cicinchen had taken an aluminum heatsink. Something like this one, just bigger:

Yes, QHD panel with G-Sync (like in the signature).

I only uv CPU core and CPU cache, more isn't necessary imo. Exactly like in the photo. Usually @ 44x you can uv much lower but for the comparison i chose the same settings.

 

For my part i'm really really satisfy of this Copper heatsink now i can game 2 hours @ 4.4 Ghz under 70c degres and my gpu around 62 with my fan speed at 75%... so 11c degres less and 8 dba less.

 

Hi guys, mine has just arrived and looks perfect!

 

Can you ask if he'd also be willing to make a heat sink for the mag 15 with heat pipes? The heatsink is good but it looks like there are too many CPU pipes shared which causes the gpu to heat up too much. It's a tonfeng chassis qc7 I believe.

Sent from my Moto Z (2) using Tapatalk

 

Send me a link for the stock one I will ask him.

 

I'll try to find it but atm it might be too new. It's tongfang QC7 if you can find it. I'm asking some owners for the heatsink serial number but I'll have to wait

Sent from my Moto Z (2) using Tapatalk

 

Without the balance of the shared heatpipes the gpu wouldnt never be adequately cooled.

 

I wonder is Cicichen has the backside VRAM heatsink for a RTX 2080 for the A51-M? I know a lot of people want them...

 

Well, I got my heatsink today. Overall, condition was good, but I did have one of the fan standoffs on the left fan (when the laptop is open laying in front of you) that was broken out of the box and the fan shroud itself was a bit damaged. It looked like that side of the box might have taken a hit. I got it straightened out more or less and got it all put together, and so far results are good. Hottest CPU core is down from 100 to 82, and GPU temps are about the same. I'm pretty happy with that given that my main issue was the CPU temps anyway. I'm curious to see how it holds up over the next month or so. Even with the stock heatsink I could get good temps right after a repaste, but after a while the paste would squeeze out due to the uneven pressure and temps would soar. Hopefully this fixes that issue.

Well, after some additional testing, the CPU temps are still doing ok, mid 80's or so tops, but the GPU is going up to 90, which is higher than before. not sure what's causing that.

 

Probably too thick thermal pads.

I Dont know if you used the pads he insttaled on the heatsink but the layout isnt good. Too thick or too thin.

 

Yeah, I just used it like it was shipped. What's the recommended/optimal layout?

 

any update as to the recommended thermal pad configs? I don't want to just start taking some off and crack a die.

 

Well, I got impatient lol. I went back in and took off the second layer of pads across the board, CPU and GPU side. I also swapped out the thicker pad for the PCH and I think one other (they're obviously thicker) with one of the thinner pads that I took off the double stacks. I was very careful to evenly tighten down the heatsink so as to not bind it up in any way that might put too much stress on anything internal.

I haven't tested it much, but just anecdotally, I went from mid 80's on the CPU and pegged at 90 on the GPU in Apex to holding steady at 79-80 on the GPU and in the 70's most of the time on the cpu. I'd get some short spikes into the low 80's and then some drops into the upper 60's as well, which seemed to be related to the fans ramping up and down. Overall, it seems much better but I'll post back with any additional results after I've used it a bit more. If anyone else has recommendations about their thermal pad layout it'd be interesting to compare notes.

 

Currently setting up mine.

Gesendet von meinem SM-G973F mit Tapatalk

 

Well, for no reason seemingly I'm back to being pegged at 90c on my GPU. Cpu temps are still tolerable, but not amazing. I'm just not sure what I'm doing wrong here. Really wish I had one of the second revision heatsinks.

 

Unfortunately mine are the same.. 89 degrees on the gpu... could anyone please share the thermal pad design with each thickness?

Gesendet von meinem SM-G973F mit Tapatalk

 

You have to measure it with trial and error. Check the imprints on all pads, if excessively deep, try to replace those with 1 size thinner. Try again, if they dont touch the component, remove the other thick ones that had a lighter imprint.

Use Arctics or Gelid pads because they are far softer than lets say fujipoly etc.

 

I understand that's what you'd have to do to get the fit 100% perfect for your individual computer, but there have been others in this thread posting much better results than hans and I are seeing and I bet if they were to share their layouts it would get us *close enough* to good to at least get into a normal temp range. There are guides on here for the stock dell heatsinks for what thickness pads to put where for optimal layout, this shouldn't be any different (save for the difference between the first and second revision of the heatsink).

 

As I have said to others, and I have learned from it by followign a guide myself.

There are various Dell heatsink revisions, they use 2 to 3 manufacturers (FCN, Sunon and CCI) and all have slightly different measurements and stock pad sizes. So no you cant just follow a guide, your heatsink might be totally different. Hell even your motherboard can be slightly different that causes an imperfect fit. For example my older ALienware 15R2 had a motherboard revision which prevented earlier revision heatsinks to fit on it correctly after Dell exchanged my motherboard. Thinking you can just go by the measurements of others is just incorrect and problematic.

 

I understand that different revisions of motherboards and heatinks could cause potential issues if you've got a different set than the person your taking measurements from. However, we have a standardized heatsink here (only 2 revisions and we know what they are) so barring a different motherboard revision that changes the surface height of the components (if it changes they layout we're screwed because this heatsink won't work at all) the fit should be similar. I'm willing to take the risk of it still being ineffective. I've torn this laptop apart half a dozen times at this point to test different fits and thicknesses in different places, and even when it appears that everything is making good contact, I still have sky high GPU temps. CPU is fine. I just am at a loss and am willing to try just about anything at this point.

 

Well guys. I received the new Cicichen revision heatsink and i used the Striker thermal pad layout picsture. But some pad was to thinn so i had to use many 1.5mm and 3mm pad. I did many trial to get best contact evrywhere and with both DIE also.

If your temp are so high on gpu there a big air gap between your DIE and the copper heatsink.

Make no sens to get 80-90c on gpu. I see you use conductonaut? You should get around 63 to 67 degres at full load.
And my cpu on stress test @ 4.5 GHz never exceed 70c degres.

Best way to look the thickness of the pad is put the thermal pad on your MOBO and dont peel the plastic from this side. Peel the plastic only on the heatsink side. Screws the heatsink and see if the pad glue to the heatsink. If the pad stay on the mobo then your pad is too thin so you a thicker one. Same thing on other side if the pad glue on first time try a thinner one. I know its alot of time but its worth it!

 

See, that's where I'm confused. The thickest pad I'm using is 1.5mm and I'm still getting sky high GPU temps. If anything, my pads should be too thin, which would mean low core temps, but possible throttling issues due to VRMs or DRAM getting too hot.

 

1. Where can i see what revision of heatsink i got?

2. Could you please share your layout?

Gesendet von meinem SM-G973F mit Tapatalk

 

The new revision heatsink have some Thicker copper plate at 3 diffrent place to allow to use thinnner pad.

Its hard to share a layout when i dont know what your heatsink look like

 

On the GPU Vram i used Arctic 1.5mm pad and with the new revision heatsink you still need 3mm thick pad on MOFOSET

So if you have the first version of the heatsink you will need 4mm pad On MOFOSET

 

Look at the Striker Layout. Most of them are good but he had the first revision heatsink. So i had to make some modification on Vram. other wise evrything was good.

 

That's what mine looks like if that helps to disclose the revision.

Gesendet von meinem SM-G973F mit Tapatalk

 

You're saying that you are using 1,5mm pads but on the picture orange says 1,0mm..

Gesendet von meinem SM-G973F mit Tapatalk

 

Hard to see.. remove the pch thermal pad

 

Hope that helps

Gesendet von meinem SM-G973F mit Tapatalk

 

That looks like the second revision to me. The two raised rectangles to the right of the GPU area as well as the raised bar on the left of the GPU area (as oriented in the pictures) are I believe the changes from v1 to v2. At least, that's my understanding.

 

well as someone else mentioned, your heatsink definitely isnt making proper contact with the GPU die. THis suggests that some pads at least are too thick. What you can do is to use a 0.5mm copper sim on the GPU die and see if your temperatures improve.

 

Yes its the new release.

 

Like I said I have the new revision 17 R5 and I used 1.5mm artic pad on Vram and 3mm On MOFOSET.