*** Official Clevo P65xSA/SE/SG / Sager NP8650/51/52 Owner´s Lounge ***

Try these: http://forum.notebookreview.com/ali...op-driver-mod-780m-880m-980m-mobile-gpus.html

 

i7-4720HQ.

 

!! THANKS FORUM PEEPS !!

After trawling this forum in even more detail (You people sure do make a lot of posts btw lol) I found a solution to all my stuttering problems! It turns out my laptop shipped with the 1.0.3 EC rather than the 1.0.4 EC available from XMG. I was hesitant at first but the difference has been night and day! I owe you all beers!

 

Reminds me a question I have. Htwingnut in his review of p650SE/P650sg report temperatures for cpu way higher in benchmarks for p650sg+4720. I assume cpu cooling systems are identical in the two laptops, so could it be that 4720 heats a lot more than 4710?

 

Hello everyone.
I have a little question:
I ordered my P705 Pro from Schenker. Everything fine so far. However they said me, that I will get the "new" Intel 4720 instead of the 4710. my Order however was the 4710. They offered me a free exchange. But it causes work in terms of deleting my personal Installed files, remove the installed SSD shipping and so on.
The question is: Do i really need the +100MHZ? Normally i would say "why not". But in terms of Overclocking within the TDP, the CPU always stays between default clock and under maximum turbo freq. So do i end up with the exact same frequency and power?

Best regards
Jubeltrubel

 

ok im a lil confused by ur post. uve already received ur machine and mysn offers you to send it in and exchange the mobo in order for you to get the new cpu? in that case just remove ur hard drive and send it in without storage. the cpu swapout shouldnt cause any problems in ur already existing installation.

if, however, they have not shipped the laptop yet (which im assuming, cuz only then it would make sense to change ur order to the 4720HQ), then how could you have installed any files on it already....???

as i said, im confused, please clarify

 

This is very unlikely. All MQ-HQ processors are basically the same. 4700, 4710, 4720, 4800, 4810 and so on are just different iterations of the same die with the multiplier increased by 1x for each stepping ( 4700@2,4 GHz, 4710@2,5, 4720 @2,6, 4800 @2,7, 4810@2,7 ...) and so on.

Probably is just a question of luck: if you get a chip from a good silicon wafer, you're going to have a cool chip, easy to overclock. On the other hand, a bad wafer may result in higher temps, throttling, and so on. This why, for example, AMD made the first 3-core CPUs... they were just 4-core CPUs with one faulty core/bad silicon. Disabling one core made the CPU usable, and more importantly, AMD could still sell them for good.

So a hot 4720 maybe is just a bad one. Temps difference with 4710 (and 4800 or 4810) should be minimal.

 

This is what I thought... And since people generally report abnormal (=high) temps we generally only notice them

Too bad it is soldered then, you have to be lucky on both cpu and gpu !

 

That's the risk of this laptop. Luckily my 4720 stays between 85° and 90° in the XTU stress test with default vcore, even if I raised the multiplier by 2x.

 

Hey there =)
Its a bit tricky for me as a german to describe the problem precise enough. However ill give it a second try
I´ve received my Laptop. My order was 4710, when i asked about the logistic delay (in the meantime i saw that schenker has only the 4720 in place, instead the 4710) i asked if i then get the 4720 or 4710 (because mine wasnt shipped at that time). The Service told me, that i will automatically get the new 4720.
So, i asked why i received the 4710 instead of the 4720 (just for my information, not that i desperately want this 4720). They offered me a replacement. I got a RMA Number and so on.

BUT, they question is: Is it worth? because this CPU has a default (non-Turbo Clock) of X. The 4720 has the Clock X+100. But this is only true in terms of default clock (no turbo active).
However every CPU these days goes into turbomode (you know the default Clock X + [depending on active cores..] X+Turbo Clock=performance.
BUT just in range of the TDP and temperature range of the CPU. We know that the CPU cant go all the time maximum turbo. So the normal usecase is always between default clock and under max Turbo. Given that its the same chip, it would then just go higher in terms of turbo boost state.
A bit of a complicated one, but i think overall in terms of gaming im not noticing the +100MHZ right? In terms of office the power is enough anyway.
Thanks

 

aaaah, gotcha! well, in the end it really depends what the main usage of ur machine is be aware that u can add additional +200 Mhz via intel xtu and thus overclock both cpus slightly. in games it wont make any difference whatsoever, but if u have other workloads such as mathematical calculations / CAD / renderings / whatever, then ull be happy about any extra mhz u can get ur hands on

 

Cpu throttle issues.

Max temp: 80 degrees
Speed when throttle: between 1,6 ghz - 2,5 ghz when playing bf4
Gpu: overklocked to 160+ core
BIOS: Premas bios/vbios
volt: no undervolting
oc: have tried bth no oc and oc. no different

Have tried: Throttle stop does not work. have tried lowereing the volst, no differens. fan is indeed spinnig decent loud.
Any ideas? would love some help.

Cheers

 

Those are some good temps. My EC is v1.03.04 and my BIOS is v1.03.05 so I don't think that is the issue.

I contacted Eurocom support and they said that mobile CPUs are rated up to 105C. Which means they probably won't do an exchange.

Although I just ran the test again after returning to default values, and stayed around 90C for most of the run.

 

A tip is to lift up the pc my one inch or so with two caps from a coke bottle or so. that way the cpu can take in more fresh air.

 

I do have to say I hope I hit the silicon lottery, little bit worried with everyone talking about really high temps on the 4720hq.
I'm used to always monitoring my temps as it is right now (I have a 920xm in a g73, wasn't meant to go in that laptop and it easily gets into the 90s when stressed).
I guess I'll just see when I get my P650SG on Monday/Tuesday


Sent from my iPhone using Tapatalk

 

With a bit of an undervolt and a repaste most people with higher temps should be able to sort it out.

 

Crap..Wrong thread..

 

Tried to undervolt the CPU and with great surprise this produced higher temps. With 20mV undervolt I go up to 95 degrees, while stock barely goes over 90.... This is quite odd.

 

If anybody's interested, listed my machine in the NBR marketplace.

http://forum.notebookreview.com/notebooks-desktops-sale/769742-fs-eurocom-m5-pro-clevo-p650se.html

Probably most appealing for Canadians wanting it at a discount. Great little machine overall, but I miss my nice, big 17" screen.

 

Yeah i RMAd my p505 because of a dead pixel. They said that they have no spare Display in stock so they exchanged the whole Unit and i got the new Core i7 4720 instead of my old 4710 and i also got an extra game code for Assasins creed. Very happy with the support of mysn.de and the extra 100Mhz.

 

Hello, where can i find the latest stock BIOS? Premamod, sager or mysn?

 

CPU cooling is not that great in these models. You'd need to repaste with some good paste and pop the back of the machine up for some temp reductions. It's not that hard to manage, and many people choose to keep auto fans too; which is a killer of the cooling system. So don't worry.

As n=1 and I have found out, slight undervolts at stock speeds apparently keep temps and such in such check that the machine keeps its max boost more frequently, opting for higher overall temperatures with better performance. You may wish to start testing undervolts at -50mV. Some chips even manage -80mV or less at stock without issues.

 

I'll probably just chuck the arctic silver 5 I have laying around on it and I would like to tweak the fan curves a bit, my current laptop doesn't let me


Sent from my iPhone using Tapatalk

 

Arctic Silver 5 will shrivel and die on these. You want one of the following:
IC Diamond
Gelid GC Extreme
Arctic Céramique 2

and the bad boy

Coolaboratory Liquid Ultra

The final paste is liquid metal, so be careful if using it. But its cooling *IS* top notch. Literally doesn't get better than that.

 

Is this lappable or is it too hot without a cooling pad? I'm so close to pulling the trigger on this but all the high temperature reports are kind of scaring me.

 

Should be ok, check the reviews from htwingnut and notebookcheck, under surface does not have crazy high temperature.

However, I assume the p650sg should be more able to lap gaming since it's underside is plastic, where the p650se is metal

 

I have a similar concern as well. I assume a coolermaster notepal u3 cooling pad would be sufficient for this model right?

Also, does anyone have experience with the Transcend 512 GB m.2 SSD (TS512GMTS800). This is one of the few aftermarket 512 ssd solutions that I could find fairly available on amazon/newegg/ebay. The Crucial 512 m.2 SSD was not readily available and when it was it was nearly double the price. I would only be using this m.2 ssd drive as storage for games and not as an operating system SSD as I was planning on having either a 850 pro in the 2.5" slot or an xp941 m.2 as the operating system drive.

Also, I see that the Corsair m550 m.2 option has been replaced on Sager's site with the Micro m600 which seems to be a slower performer than the Corsair. Since the Transcend drive is cheaper, it seemed like it might be a good potential option for $75 cheaper than the Micron model.

 

I used CLU... I guess I should do some tests today. Been playing too much CSGO and FFXIV (both games cap around 65C at their respective max settings but they arent stressing). Guess I should just run that BF4. I have no undervolt either.

 

That would be great LoneSyndal...I'd like to hear your experience after playing BF4 since that seems to be the best "real world" gaming application for testing these new laptops.

 

The chassis itself doesn't get too hot, so I wouldn't worry. The GPU remains relatively cool (haven't seen it above 70C) - it really is the CPU that's an issue.

I don't understand modern laptop design. They put big (and multiple) fans on the GPU, but the CPU is the one that needs extra help cooling itself.

I also have tried using a cooling pad. I didn't see any appreciable difference.

I don't think air intake is the problem here. It's small fans combined with a small heatsink.

 

Here's info:

Without Turboboost:
http://puu.sh/eVgBT/4f99ec3152.png

With Turboboost:
http://puu.sh/eVgEv/30fb908be7.png

You will see that its throttling from current limit and nothing else. No undervolt.

EDIT: BF4 on Wake Island 64-man server.

 

Do me a favour. Please set the following settings in XTU and tell me if you keep your max turbo in BF4.
Change "Turbo boost Short Power Max" to 70W.
Change "Turbo Boost Power Max" to 70W.
Change "Processor Current Limit" to 100A.
Change "Turbo Boost Power Time Window" to 128 (seconds).

Let me know if you keep max boost in BF4. Make sure your power plan sets your minimum processor state to "100%" (I think High performance does this by default).

 

http://puu.sh/eVp9P/80af9f0e3d.png

Doesnt keep the boost settings. Current limit still there.

BF4 Wake Island 64-man server on Ultra Preset 1080P settings.

 

What do you mean it doesn't keep the boost settings? When you "apply" in XTU, it reverts to default? Are you using Prema's BIOS for the machine?

Or do you mean it doesn't keep the boost CLOCKS and throttles even with the power limits supposedly relaxed? If this is the case, the HQ chips are entirely pointless and overclocking them is a joke.

 

Pretty much it cant keep the boost clocks. Guess that current throttling is the last hurdle.

 

So the chip is a piece of junk. Glorious! You're the second user who has confirmed for me that those HQ chips are worthless and cannot keep their boost clocks under load. Even under stock settings for my 4800MQ I don't drop clocks under that kind of load. If I really ream it in some applications, it'll demand more power for sure, but usually it doesn't pass its inborne limits for even BF4 gaming.

 

Um... so, you are telling me that a 47W CPU needs more cooling than a 125W GPU?

I'm not a mathematician, but AFAIK 125 is bigger than 47...

 

Haswell CPUs get hotter than the GPUs this time around, I think. They're insane. Maxwell GPUs are EXTREMELY COOL. 980M is only 100W and 970M is only about 80W. Also, under a good bit of load, the mobile CPUs draw as much as the desktops do. So they can hit 80W, 94W, etc. Here's some proof:
80W at stock speeds and -50mV undervolt
94W at 3.8GHz at stock voltage

 

Re-ran the settings using 3D Mark. Didn't get a higher score due to not being able to overclock further on 180W, giving me power issues if I tick settings higher with CPU running at higher wattage.

http://puu.sh/eVGoj/62a710df02.png

 

... I was expecting you to run your GPU at stock. Maxwell is a power drinker when OC'd. Were you testing your CPU's boost with an OC'd 970M? That's a healthy OC you have on it, and thus not nVidia's stock vBIOS.

 

I usually run with slightly less memory overclock on the 980M. I can do a quick run with CPU adjustments only.

Here's a Prime95 run. I added a -50mV undervolt during middle of test, thus showing the drop to 93C at the latter half of the run.

Max TDP was 56, consistently running at 3.19 GHz. No other throttling besides Current Limit.

http://puu.sh/eVIQh/dedc8a01be.png

 

Oh wow you have so few CPU controls... I feel sorry for ya. And yeah, your CPU isn't pulling any extra power even though you set it to. So those chips ARE in fact TDP locked no matter what. That's wonderful. Thanks for the in-depth testing, man.

 

That would be the compromise with going for a thinner machine. Especially when being on-the-go. Luckily for me, I do have a buyer for the P150EM which I have modded for GTX 980M support (at their request). Should be able to get myself Batman as a good DTR for my workstations at home (much cheaper too as I already have the M.2 M550 SSD 512GB, i7-4790K delidded, 16GB RAM, and more...). Though my purposes for the P650SG isn't entirely for gaming/performance as I have stated earlier on in thread, work comes first.

 

I understand. It's just... the 47xxHQ series also falls into the 4980HQ line that the GT80 and such are coming with. If those chips follow suit, then even my TERRIBLE silicon-lottery'd 4800MQ is better.

It's pointless even marketing a machine as high end or boasting that you can OC a TDP locked chip.

 

Indeed. I am quite disappointed with most chips after the Ivy Bridge lineup. Guess thats Intel's monopoly for ya.

 

Um... these screenshots don't prove anything. Software calculated power consumption is just a heuristic guess made by a software that can't actually measure anything. I'd believe those values if they were measured at the plug...

 

Then don't believe me. All I know is if I try to livestream with my CPU clocked at 3.9GHz and 112A current and 120W power allowance, I will shut off due to overloading my power brick, even with my cards at stock, due to the excess power draw.

He used HWMonitor and his chip barely passed 50W.

I JUST ran Linpack (which was a gimped run because I had a large amount of programs open at the same time) and I pulled 83W. It'd pull even more if I closed down all the CPU-eating background programs like Chrome/Teamspeak/etc.

Spoiler

 

WOOOOOOO!!!!!!!!!!!!!!!!! Phase 2 from XoticPC (QAQC Phase).

Was so worried I was going to be left in the dark over the weekend, this is excellent news, as I hope to get it in my hands before next weekend. I have a friend who will be in town who is interested in the laptop. If he is impressed, I will direct him over to Xotic to order, so they may have made another sale.

Hope to see shipping early next week.

 

Actually, I think there is a fair amount of difference. I just got my NP8651 yesterday and have been putting it through its paces. I tried running IXTU CPU stress test with the multiplier at 4700, 4710, and 4720 speeds. Each bump in speed produced a 5-10 degree average difference in temps, with the largest difference between 4710 and 4720. At 4700 speed, temps stayed mostly in the upper-70s. At 4720 speed, high-80s to mid-90s. That's a pretty big difference.



Of course, the other explanation is that I got a bad chip, I guess. But this also happens with desktop haswells - each 100 MHz increase can result in pretty significant temp increases.

 

What paste do you have on your 4720HQ?