17" R5 Just How Bad?

This is a run of yesterday evening.

I get your point, but if you want real performance in a laptop form-factor ou have to pay man.. there is no way in hell BGA will run as pure as LGA that's why BGA is cheaper. They're selling is broken chips repurposed as new chips (just because it got a different name)

LGA is the way to go.

 

Well to be fair, the one you referenced was 5.087 Ghz. Also, maybe the person you linked to was sure to turn off everything in the background, changed process priority, etc. Can't really see the Linus screen that well either, maybe it had a blip during the run down to 4.9, although they seemed to indicate it didn't throttle. What surprises me is the fellow in the link used liquid cooling and all Linus did was use forced air...

 

that 0.087 won't make a difference, and he did that because the 8950HK is not stable on air ! that man LM'ed his machine, and uses watercooling cause with air you're gonna have alot of trouble (BGA) getting it stable with around 1.3/1.4 Volts. Not only did he not turn off nothing in the background, he even left his AV on proving that he got it stable in a normal condition (except that the laptop was till bareboned) and that he lives dangerously

 

Thanks for the screenshot. I'd be happy with that! I get what you're saying though and I agree. I was all set to return this and go LGA, but for various reasons I want to give this one a last shot before I do that. I have another week anyway to decide. To be honest, it would be difficult for me to wait another three weeks (for it to be built), then the 9900k is coming out soon so I'd wait for that before even ordering, etc. I've already done 20 hours of downloads/installation and have another 20-30 hours of installation to do. (lots of audio stuff doesn't allow for local installers unfortunately and can't be cloned over to a new system, has to be re-installed) Plus, I really like this machine in a lot of other ways - per-key lighting, love the keyboard, trackpad, IR camera, sound, and overall design. LGA is going to be 20% heavier, thicker, even less, a lot less, battery life, no per-key lighting, no IR camera. Even if the LGA cost the same as mine, it wouldn't be such an immediate decision to switch if this re-paste works. I'm in so deep now, have the thermal materials on the way, etc. Might as well see how it goes.

I do appreciate your concern and everyone else's. Like I said, I had an order worked up and ready to go, but after I sat on it overnight, the sticker shock was just too much. What should actually be the cheaper option (thick case, basic design) ends up being much more expensive since it's not mass market.

At some point, I wish this concept had come to fruition, that would settle it, we could just build our own and have the best and cheapest option:
http://www.binari.tech/

 

You're going by the screenshot I take it? His Wi-Fi is not connected, I noticed. Anyway, not sure why the difference. Your own score is 60 points from where you originally said though (and I've seen mine fluctuate within a 50 point range easily), so I don't see why this other guy getting 100 points higher with a slightly higher speed isn't possible...

 

I will also add - I think the Asus model that was in the video is very overpriced. I forgot what it was, I think almost $4k? Obviously at that price level, LGA becomes the obvious choice. But at the price these Alienwares come in when you talk to the sales rep and work out a deal, they're pretty good, especially given the feature set beyond the basics of cpu/gpu speeds.

 

well the difference between my and that guy, is that i have to play with the priority settings before it score's the way i want. + i have to put it on High performance mode, putt speedshift epp to 0 and turn off AV.

This was a run done with high priority settings

 

Wow that is quite a difference. I hadn't researched how much a difference the priority setting makes, but it is (I never bothered changing it myself to test). I always run high performance and epp at 0, although I don't disable antivirus for the testing. I see what you mean then, the fellow you linked to isn't doing any of that. I guess at some point, some folks are going for the high score, but I just use it as a general measure of how much CPU capacity I can expect to have to load more media/plugins/etc. I wonder what the fellow you linked could have gotten if he'd optimized for the benchmark!

 

Blip down to 4.9GHz? Doesn't matter. 4.9GHz is still +1640cb.

Wrong! Liquid Nitrogen and the MB out of the chassis. Liquid Nitrogen is Sub Zero temp. Often -90C.

Could you put High Performance Power plan so clocks stay where they should? Aka NO drop.

 

Yeah that's what i said, it's bareboned. maybe i should've said it clearer. That run was a high performance powerplan run without dips. All my OC values are with a speedshift epp to 0

Edit: I know man all my score's are too low BGA curse continues from my R4 to the R5..

 

I never see this numbers.

 

Yeah that's because when Cinebench is on top, it overlaps HWinfo

Edit:

Here is a run brother:


High performance mode, Speedshift EPP - 0, High priority in TaskMgr. BTW i'm using (chrome) remote desktop so it can lower my points but does not necessarily have to (my screen locks up until my laptop is done with the benchmark)

 

Does anyone know if the same heatsink is used on the i9 with the 1080 GPU and the i7-8750H with the 1060? I have ordered the i7-8750H with the 1060 and I was wondering if I could install a more capable heatsink and avoid a possible overheating issue and because it seems to be a PIA to repaste I would like to replace the heatsink at the same time as I do the repaste.

 

Why don’t you just wait and see how your laptop works and go from there....

That setup puts out a fraction of the heat the i9/1080 combo does. If you read through the history it’s not the folks with the 1060 having all the issues.

Don’t create a problem where none exists.


Sent from my iPhone using Tapatalk

 

Waiting would be okay except I'm only going to be home for a couple of weeks and then away for about 6 months. My plan would be to have all the information at my fingertips including where I can buy the better heatsink. As soon as I get my new computer I'll do a heat / stress test and if I am having problems then I'll buy the better heatsink and install it when I do the repaste. That is the plan anyway.

 

I doubt they make a different heat sink - this one shows the heat sink from R4 with different 1070 vs 1080 and CPU choices, they all look the same as mine which is new (i9/1080) (this is the iUnlock repaste guide, I am sure you saw it already):

http://forum.notebookreview.com/thr...5r3-disassembly-repaste-guide-results.797373/

If they didn't change it across model years, then that means they certainly didn't upgrade it for the i9 (which they should have).

Like cruisin said, I highly doubt you will have any issue with that combo. I don't even know I've seen any threads of people repasting with that cpu/gpu combo. The 8750H is truly the ideal match for the heat sink and overall platform size/design. I mean, look at the difference in cores - when I run a one core test at 4.4 Ghz, I think my max temps are not even 60. Whereas on six cores I am fighting to keep it under 93, even after my re-paste/re-pad. Basically, if I limited my chip to 3.7 Ghz before re-paste, it ran fine, just hot (but no throttle). Being that I think 3.7 is the 8750H's max stock turbo, I don't see how it could have an issue.

If you do want to improve efficiency, the re-paste is going to do all that anyway (and maybe changing the pads will help); I don't really see what else they could have changed on the heat sink itself, unless somehow yours comes with less pipes, but that just wouldn't make sense, they all have to fit to the same motherboard and chassis anyway.

 

I had seen the re-paste guide but then went looking for it earlier to compare what I was looking at on ebay and couldn't find it so thanks for the link. I know in the past that Alienware and Clevo have used a different heatsink for the top configurations due to higher thermals and in the two cases that I remember it was just an additional pipe compared to on the lower configurations. I assumed that it might be the same here. I guess I'll wait the couple of weeks until I get the computer in my hands and then if I do have to open it up and do the repaste I'll compare what is in there to @inulock photos.