Asus ROG GL702ZC owners lounge

You wrote somewhere before that both CPU and GPU were in acceptable thermal level and fans just get suddenly to full speed.
This is proving that you are wrong about this. CPU and GPU are cooled GOOD even without undervolting on this cooling solution.
They temperatures were in acceptable level and fans had a lot of space to speed up if temerature of CPU or GPU would rise up.
This is only B350 that is cooled unadequate and fan controller is blind for its temperature.

I'm still trying to say undervolting is good, but not for this specific model with old cooling system.
It will be really good with new cooling design.

It is full unmodified DESKTOP class B350 chipset inside laptop... WITHOUT temperature dependent active cooling system... even without proper passive cooling.
Look at B350 MoBos pictures... how big the passive radiator on B350 is. They preasumpt that these fans will be spinning to cool down B350 too.
Only new cooling type with heat-pipe on chipset can be adequate to cool it down - in this solution B350 temperature is measured INDIRECTLY by GPU temperature sensor.

Take wall power meter and prove I'm wrong.
I'm continuously connect my laptop to wall power meter since i bough it.
I even know that my laptop fried when power draw from wall was around 50W and when fans bump up to full speed power draw increase to 65-70W (so they are really power hungry).
I can't check it right now by myself... I have to wait till they return my laptop.
I'll check it in real conditions as soon as I get it back.

You have partially right... fans are spinning depending on two temperatures, CPU fan depends on CPU temperature, GPU fan depends on GPU tempetarure, but NO fan depends on B350 temperature (*on old cooling system design).

But think about it differently. You are cooling one part of system better (actually you are heating it less not cooling better). And this part is forcing that the rest is hotter. Just because the rest can't control fan speed. Its not only B350, its NVMe, HDD, RAM... all these elements will be hotter just because fans generally are running slower and these fans do not know that rest of the system is too hot. They only have informations about CPU and GPU which are OK.

Did you check how 5930G cooling design looks like? Did it have unmodified DESKTOP class nearly 6W TDP chipset without active cooling solution?
I checked some pictures. In 5930G chipset is under radiator with heat-pipe - this is good thermal solution and fans will spinn faster if chipset temperature goes up - like in modified thermal solution in GL702ZC

I dissagree - they would be on periodically to cool down CPU or GPU, not whole system. Fan controller don't know any temperature except CPU and GPU.

Take of bottom cover, turn laptop 90degrees to LCD be on table and do nothing on windows without internet conectivity. Touch B350 and prove I'm wrong. You can even cool down / heat up it and there will be no difference in fan control.

This is only my theory. I'll check it in real conditions with good quality thermal camera when they repair my laptop.
One thing is that this metal plate is silver - it will do some reflections. I tried to measure temperature of this plate with touchless thermometer but it wasn't working because IR reflections.

You are right, it will never go above balance point (which depends general on power dissapation on all components). Balance point depends on enviroment temperature and actual power draw of components. Thats why I'm simulating constant power conditions with windows iddle and no internet connection (to not allow programs to do something in background).
The thing is in these conditions fans will do they work only for CPU and GPU cooling. Balance point for B350 will be a lot upper balance point of CPU and GPU (which is controlled directly throu fans).

If it's OK, why are you so concern about additional radiator .

Don't get me wrong. I'm without laptop and can't prove my theory is true. But you have it and you can prove my theory is wrong.
I'll be glad if you make some tests. But I don't know if you have manual skills. I'm PCB maker and embedded programmer I have skills and acces to equipment.
My theory is directly related to other systems that I have/had and saw and few tests I did witch mine GL702ZC.
If my colleague agree I'll make some tests on his GL702ZC at work.

Once again we are discussing THEORY not real conditions. Like politicians Almost no real life knowledge

 

When I was playing Mass Effect Andromeda, I did NOT have the GPU or the CPU undervolted - you keep missing this point - I was playing the game with an UNMODIFIED system.

Another point was that when the fans decided to kick in to 100% suddenly, the temperatures for the system were the same as they were in other instances of playing mass effect Andromeda.
ME Andromeda stressed the GPU so it was usually hitting about 78 degrees C at the time. To me this doesn't indicate a dangerous level for the GPU or the CPU... but you are likely correct that it's not too good for the B350 chipset which doesn't have proper cooling attached to it.

Undervolting played no part in the problem I developed.

And I keep telling you that undervolting likely has little or nothing to do with it because my problem happened BEFORE I even attempted to undervolt my system.
Undervolting only kicks in once the system is STRESSED and it reduces the power draw of two major components that usually spread the heat around the system anyway (the hotter the heat pipes get, the hotter the entire system gets... and fans are responsible for cycling cool air into the system - when the temperatures on the CPU and GPU are on a certain level, the fans will usually stay at that level anyway.

Also, the cooling design that has a heat pipe running on top of the B350 chip is not new... and there's nothing to confirm whether its actually 'good'.
That's a design that ships standard with GL702ZC that has Ryzen 1600 cpu inside - plus the operational temperatures of that system were slightly higher under that cooling than for the 1700 system as reported in some reviews.

Furthermore, do we have proof that the B350 chip which doesn't have a cooling pipe running on top of it is responsible for the problems we all experienced?
Mind you it's a good indication that its part of the problem, but didn't people's GL702ZC's also failed while having that pipe there in the first place?

When another user reported the same issue with the fans like I had on his GL702ZC, he also had an additional problem... the GPU was fried/unresponsive and the LCD didn't work properly.

My GPU and LCD were running fine (that I know of)... I was still able to run and play Mass Effect Andromeda, but I didn't want to continue using the system with fans working at 100% all the time as I knew that 'something' broke in the system.

The CPU and GPU on the GL702ZC are too close together.
That's a design flaw inside the 1700 and 1600 versions because they are identical in motherboard layout, except which CPU's they have inside.

I think the cooling system for the GL702ZC needs to be redesigned ENTIRELY... not just for 1700 version, but also for 1600.
And I don't think Asus has been working on anything like that.

If some people had their cooling assembly changed after RMA-ing the unit, I'd like to see whether the 1700 version received the cooling assembly without the pipe running atop of the RAM slots and has 1 connected to the B350 - and it that's the case, then that means that Asus simply decided to use 1600 cooling assembly in the 1700 version.

When have I claimed that the B350 dictates whether the fans will cycle up or down?
I merely stated that undervolting drops overall temperatures generated by the CPU and GPU when they are STRESSED, which also drops the amount of heat that's present in the system because those 2 components contribute to overall temperatures inside the WHOLE laptop.

I know what you're saying, but I don't think you have a particularly strong case here.
You could say the same thing for other laptops whose active cooling doesn't cover the HDD, RAM, NVME... which is practically ALL of them, and yet they don't seem to be failing all the time (and this is talking about systems whose fans don't seem to have particularly effective or strong fans).

But, we can actually test this under practical conditions and compare temperatures of the HDD and NVME with and without undervolting applied - and I'll separate some time to do this (plus, some others here can do the same thing).

[/quote]
Did you check how 5930G cooling design looks like? Did it have unmodified DESKTOP class nearly 6W TDP chipset without active cooling solution?
I checked some pictures. In 5930G chipset is under radiator with heat-pipe - this is good thermal solution and fans will spinn faster if chipset temperature goes up - like in modified thermal solution in GL702ZC
[/quote]

Wrong.
The 5930G only has 1 fan (which isn't particularly loud or powerful), while the CPU and GPU share the same cooling. This isn't a particularly good design, because the amount of heat is still higher when you have both the GPU and CPU sharing one cooling system and only 1 fan (makes the system more difficult to keep cool and prevent thermal throttling). It's better to have the CPU and GPU each on separate cooling with another fan in that case.
When I undervolted the CPU and GPU on 5930G, it also dropped temperatures inside the entire system (despite the internal fan running lower).

Also, when the 5930G was stressed to the maximum (without undervolting), it's fan wasn't particularly loud to begin with.

One of the problems we have with GL702ZC is we cannot measure the chipset temperatures on the B350 because none of the software apparently reports it.

You are correct that the fans only respond to the CPU and GPU, but (and this is the third time I am saying this) they are also responsible for cycling the cool air through the system.
If the general system temperature inside the system rose, in itself, it wouldn't affect the fans, but it would affect the CPU and GPU temperatures (which would then prompt the fans to power up and cool things down).

You do realize that touching the motherboard while the system is active runs the risk of frying it in the first place?
Why do you think its recommended to shut off the computer entirely and drain the integrated battery (not to mention press the power button after all that to remove any excess charge in the system) before you start doing anything on the inside?

Sorry, but I won't be touching the motherboard at all with my fingers while the system is active.

We are spinning in circles, and I don't wish to repeat myself again.

Because I never said that undervolting is a be all, end all solution.
The B350 chipset (if that's really it) not being connected to the rest of the system via a cooling pipe is plain stupid, and having a radiator there would be helpful.
My point was that undervolting helps the overall system (not just the CPU and GPU) stay cooler.
CPU and GPU both affect temperatures in the entire system, and, by dropping their power consumption and temperatures, you also affect the temperatures in the rest of the system.

Well, I'm not a PCB maker nor an embedded programmer, I'm simply reporting my experience based on my own observations.
And those observations told me that undervolting was NOT responsible for breaking my system the first time because I didn't even have the system undervolted.

My subsequent use of the laptop (after the repairs) consistently show me that running the system with the undervolt on the CPU and GPU reduces overall system temperatures as well (because the keyboard area you mention is cooler with an undervolt being active) despite the fact the fans don't run as hard as they do without undervolting (because, without the undervolt, the fans would simply cycle down when the internal temperatures of the CPU and GPU reach lower levels anyway - so having the temperatures at lower levels in the first place just keeps the fans running on appropriate levels to begin with).

My system hasn't failed yet since running undervolt on the GPU and CPU... but, I shall test the HDD and NVME temperatures with and without undervolting...

I still maintain the GL702ZC needs an overhaul of it's entire cooling system. But the fact the CPU and GPU are so close together on the mobo doesn't help things at all

 

Could you post idle and load temp on NVMe drives if you have it in your system. Boiling NVMe ssd temps as in the Alienwares will as well mean high chipset temp. Several NVMe drives in the Alienwares show fast decrease in drive remaing life due the heat.

 

Good evening to all, just bought this laptop at it's GC187T version (main differences the 8GB RAM in one stick and the one HDD of 1TB storage). So i am planning two steps of upgrade.

1)I am adding a Samsung 970 Pro NVMe M.2 512GB. Is there any detailed guide for how am i gonna execute this and make the samsung my boot drive?

2)Then i am gonna go for more RAM. Should i buy another 8GB RAM like the one i already have, or should i take it out and buy 1 16GB stick of RAM. I cant afford upgrading with 2 new 16GB RAM sticks at the same time. What will be the difference in performance (mostly in Premiere Pro, no gaming here) between 16GB (2x8) dual channel and 16GB (1x16) non dual channel? Is it worth it to make this compromise and wait to add another one later?

 

About RAM. It's always better dual mode I'd say. But if you are planning to add 16gb later (to get 32gb and if you need it) - I think your better choice to find 16gb stick with high clock rates. 2400 is good but for Ryzen it's can be better. PERSONALLY I take second 8gb stick (same 2400 Samsung) for now. But I'm not a pro user (working on Adobe/Corel too but not often and not for work). If you working on laptop for money - you better take good RAM. If you casual user - you can go for second 8gb stick and forgot about it for some time with 16gb dual mode.

 

Got this e-mail from ASUS CS recently when I tried to look into this:

No idea if AMD is planning on updating it but we'll see what happens. I'd love to have this laptop be able to upgrade its CPU, even if it means using something like Prema to get it to work.

 

Good point. That's why I recommend using M.2 SATA drives, they run cooler (test a few to find the lowest temp current models) than NVME and cost 1/2 as much - look for a sale too to save more.

I worry about the NVME drives running in laptops with circumvented cooling too, those holes punched in the bottom case right next to the heat exchanger exhaust fans.

Those holes cut in the case cut's off / reduce's the air pulled in over the motherboard / storage in the front, which is exacerbating the heat issue with NVME and motherboard chipset's - and any other other hot spots depending on that front to back air flow too.

Even the cooler running M.2 SATA drives could probably use a heat sink on their controller - or heat spreader across the whole card spreading the heat radiant over a wider area.

 

Have been out. Fired up my machine now but as always, NVMe drives in my laptop haven't problem with heat. And I have removed the black bottom lid shields long time ago. Doesn't make the big differences in ssd temp. Of course it depends on how thick/big the notebook chassis is and what types brand drives you have. The trash Dell use will probably run a bit hotter in this chassis as well. But NVMe in thin and flimsy ain't the best or smart. They will run hot and heat up the whole hardware. One of the reasons the MSI GS65 keyboard is damn hot... The drives is smacked below the MB and the keyboard and the chassis is very thin.

 

I can do that, but I'll need to spare the time to stress the CPU and GPU collectively and see how the temps are affected.
But if I'm not mistaken, my GL720ZC has M.2 inside already.
Yup... just rechecked the official specs, and they say: 256GB SATA3 M.2 SSD

The regular operational temp of my M2 sata drive is 42 degrees C.
For the HDD it's 39 degrees C.
Don't think these temps change much under heavy loads... especially with the fans active... but as I said, I'll need to stress the CPU and GPU together for extended period of time with and without undervolt to get an accurate reading.

Remember that the CPU fan in GL702ZC does not cycle up until its temp. climbs up to 90 degrees C (if I stress the CPU only).
I don't understand why Asus would go along with this option. Why not decide to have the CPU fan to cycle up at say past 50 or 55 degrees C?

 

You have the Sata3 M.2 ssd's, sir. Those won't run hot. I talk more about the NVMe drives (some brand run hotter than others as well). They will put a lot of heat even idle. If the fan speed is determined by Cpu or GPU temp they won't run if temp is cold. In small chassis will the NVMe drives often up the chipset temp, and no fans run to cool it due low Cpu/Gpu temp.

Exactly the way Alienwares engineers prefer. I can't use my oc'd i7-4930mx in my Aw17 without use of Hwinfo fan control.
They are more concerned that people complain about noise than someone complains about too high temperatures. They rely on Intel / Amd's throttling algorithm and in worse case forced therm trip if temp being too high. Many don't even know their expencive hardware throttling. Only BSOD force a phone to tech support.

 

If I'm not mistaken, my old Acer 5903G fan used to be quiet below 50 degrees C.
Upon reaching or surpassing 50 deg C, the fan would cycle up.
But as I mentioned before, the 5930G only has 1 fan... and undervolting usually helped the system reduce load temperatures from 85 deg C on CPU and GPU to about 75 degrees C.

 

If the undervolt force fans to run less often due undervolt in load and chipset and ssd's already struggle with high temp it won't be any better. Rather worse. All depends on how high chipset/ssd temps is from before the undervolt. Have seen more than enough of chipset temp around 90C and ssd's in high 60 or ow 70' s with not very high load. Not ideal. Todays MB pcb are so small due chassis thinnes that they can't implement cooling on chipset or ssd's. No places to put screws for the cooling plate for PCH chips etc

 

Don't try to paint undervolt in a bad light - undervolt making fans run less... that's a silly twist of negative spin on a positive result.

undervolting, Re-pasting, and other methods of reducing CPU / GPU thermals if effective will all reduce fan speed and fan run-times, and that's a good thing - up to the point of reducing temperature below thermal throttling.

You may be on to something in the case of reducing temperatures further than necessary to stop thermal throttling by using LM and other extreme cooling measures. But even in those cases, tuning the fan curve to meet motherboard / storage cooling needs can be done.

Under IO load the increased CPU thermals should cause the fans to kick in, if not then adjust the CPU fan curve.

You can tune the CPU fan curve to kick in earlier - at lower temperatures - and run at higher RPM as needed to bring back more fan air movement to keep the other laptop components cool.

As long as you don't Frankenstein the case with holes under the heat exchangers and exhaust fans you'll still get enough cooling air flow to keep the rest of the motherboard components and storage cool when the fans run.

 

The fans don't run less on an undervolt, they just run on a lower speed for long periods of time under load because the temperatures are not as high as before.
The undervolt mainly affects load temperatures.

Incidentally, the CPU and GPU contribute to overall temperatures (chipset included) inside the laptop as a whole...
Also, once temperatures have been lowered by the fans to a certain level (without an undervolt), they will simply cycle down anyway.

An undervolt really shouldn't affect the chipset for the worse... if anything, it might actually help it because you lowered the temps on the most heat producing hardware in the system and probably alleviated the extra strain OFF the chipset.

Lets' say you run a system without an undervolt.
Your load temps reach about 85 degrees C on CPU and GPU with fans running at maximum to keep those temperatures.
That heat translates to the rest of the system and overall ambient temperature inside the chassis would be at a certain (higher level as a result).
If the fans manage to cool the CPU and GPU to 80 degrees C (depending on how much load they are on), they will cycle down (and cool the system less efficiently per your analogy) until the temps reach 85 deg C again and they cycle back up.

Now, with an undervolt, you reduce load temps on the CPU and GPU to 75 degrees C on CPU and GPU with fans being at a certain (less noisy) RPM level which is enough to keep the temperatures there
Under these circumstances, the undervolt will result in a REDUCED ambient temperature in the chassis... which means less of a thermal strain on the chipset as well.

Basically, you dropped the temperature strain on the overall system by quite a bit (inlcuding the chipset).
It doesn't mean the fans would be less efficient at removing the air through the system, because they will still be running at a high enough level to meet the new lower temperature threshold across the chassis.

Now, if the overall design of the cooling unit is poorly designed to begin with, that's a different story, but I have YET to see an undervolt affect the chipset and other components like the SSD or HDD - otherwise, systems with poor cooling after an undervolt (or high powered systems with an undervolt) would experience much higher failure rates.

But as I already established earlier, my problem that resulted in RMA-ing the laptop was not a product of an undervolt, because I hadn't been running the undervolt when it happened.
It was a product of a defunct cooling assembly (most likely).
If I was running an undervolt back then on the GPU and CPU, it is possible I could have avoided the problem - though, in a way its good it happened, because it exposed poor cooling on behalf of Asus).

Also, the problem as you suggest would be focused on the SSD and/or HDD... but neither was affected.
Instead, a control circuit on the motherboard which regulates the fans was what got 'fried' or 'broken'... this might have been the B350 chipset, but if it was, I would think the problem would result in a more serious issue that would essentially damage the motherboard and prevent the hardware from POST-ing or booting into OS.

 

What Papusan and the other poster might be forgetting is that CPU and GPU thermals affect the overall temperatures inside the chassis as a whole.
If you reduce the temperatures on the CPU and GPU via undervolt, you will also reduce the overall ambient temperature in the chassis and drop the thermal 'pressure' on the chipset, SSD and HDD.

Granted, I could be mistaken, but in my experience, I hadn't seen an undervolt that resulted in higher operational temperatures of the chipset, HDD or SSD for that matter.

I can understand the logic of the lower running fans not cycling enough air through the system in general, but this doesn't translate at all when you think about the premise that by undervolting, you DROP the ambient temperature of the entire system as well (therefore, there is no need for the fans to run at higher/noisier settings) because the system's ambient temperature dropped substantially, and fans cycle up and down anyway depending on CPU/GPU temperatures to begin with.

If there isn't enough air cycling through the system in the first place, that's a bad cooling design... but, in that case, undervolting would be CRITICAL to help the system function better/cooler because the ambient temperature would drop as well.

The suggestion to cover the B350 chipset however with an aluminum radiator is a great one though... it wouldn't hurt to provide extra 'protection' just in case it is at a higher risk of failing in systems with a cooling assembly that don't have a pipe running on the chipset.
It is obvious that Asus decided to extend a cooling pipe to it in the GL702ZC with 1600 cpu... but that doesn't mean that this is automatically better. GL702ZC's with that cooling design failed as well if I'm not mistaken (do correct me if I'm wrong).

 

After sometime of using laptop, I have some more questions
1. I know that you can insert 2nd ram w/o removing heat pipe, is it possible as well to pull out ram which was inside from factory and replace it?
2. Did anyone test RAM 2400 mhz 14-14-14-29? Will it work on this timings, or it will perform on lower ones?
3. 960/970 evo ssd, don't have much throttle in this laptop?

 

@Deks can you please tell how you install drivers?
First install oem drivers from Asus and then express install of chipset drivers or video drivers from AMD?
Do you install something else? Do we even need new version of video driver?

 

That's what I did before. I already had oem drivers installed, so I just downloaded latest gpu and chipset drivers and ran them separate of each other via express install.

These days, I tend to upgrade the drivers the same way... Via express install method, but when a massive windows 10 update messed up the drivers, I had to remove them completely and install them from scratch (I used latest drivers from AMD to install both gpu and chipset a new without installing oem drivers first).

 

The chipset doesn't control fans, they simply handle I/O - USB, SATA, PCIe etc, as well as storage related tech such as RAID. Based on what I've observed in other notebooks over the years, this is more likely a case of the EC needing a reset or something similar. Obviously it's impossible to know for sure, but I'd say the chipset was probably fine. They run pretty hot sometimes - depending on the TDP and what cooling is available. I remember the northbridge (before they were integrated into the CPU die) on my 990FX motherboard would regularly hit 90c and that ran fine for years. If they get too hot, they turn off.

 

Support For

• Windows®10 April 2018 Update
• AMD Ryzen™ Desktop Processors with Radeon™ Vega Graphics
• Ancestors Legacy™
  • Up to 6% faster performance using Radeon Software Adrenalin Edition 18.5.1 on the Radeon™ RX Vega 56 (8GB) graphics card than with Radeon Software Adrenalin Edition 18.4.1 at 1920x1080 (1080p).1
  • Up to 13% faster performance using Radeon Software Adrenalin Edition 18.5.1 on the Radeon™ RX 580 (8GB) graphics card than with Radeon Software Adrenalin Edition 18.4.1 at 1920x1080 (1080p).2
• Microsoft® PlayReady®3.0
  • Supported on Radeon RX 400 Series and Radeon RX 500 Series desktop graphics products.

​

 

Tnx for the heads up.
Installing just now.


Nice of AMD to continue support for RX 580... though, that's for 8GB gpu's... ours is 4GB... we might experience 'some' performance gains still.

 

potential buyer: How well does it play with Ubuntu 18.04 and is there any place in the states that has a deal going on for it? Are there plans for an updated Ryzen+ cores?

I was looking at getting the Acer Nitro5 An515-42 but according to a post today on their forums it will only be released in the south asian market...

 

I was hoping someone could test this out a before and after, as I currently own no game to pass a judgement on this, if it somewhat affects our scaled down version of RX 580

Sent from my MI 5s Plus using Tapatalk

 

In case any of you were wondering, 15mm 2.5" HDDs do not fit in this laptop. You can connect the HDD to the sata connector but it'll bend the motherboard slightly. 12.5mm 2.5" is largest you can fit properly.

Works great on Linux. Depending on which distro / desktop you use there may be screen tearing due to their shipped xorg config.

E.g. with XFCE or xubuntu edit /usr/share/X11/xorg.conf.d/10-amdgpu.conf and add TearFree support.

 

Welp I lost Freesync..... Windows 10 1803 got installed and I was getting all these crashes and stutters whenever I boot... So I tried to update drivers through Radeon Settings, then suddenly BSOD. So I went into safe mode and did DDU. Then I installed 18.5.1 fresh and everything started working normal again, but I lost Freesync. Somewhere in this thread an user suggested to change the driver from "Radeon RX 580 Series" to "Radeon (TM) RX580". I did that but still no Freesync.

Please help

 

Uninstall everything with DDU, install driver's from Asus site and then upgrade driver's with express install

Sent from my MI 5s Plus using Tapatalk

 

FML. I installed the drivers on ASUS website. When it booted there was a hiccup and the radeon settings issued a warning that my system has recovered from a hardware malfunction. I restarted and now I am getting constant BSODs, just like what I had when the Windows updated to 1803. Now I have no other go other than installing 18.5.1 without Freesync.

 

Installed Windows 7 finally. What driver do I need to enable freesync? Trick with manually choosing RX 580 (TM) in device manager don't work since ASUS drivers for win10 only...
Maybe some old drivers from AMD?

 

You could revert back to 1709 and install Asus driver's then upgrade window's and later on Radeon driver's

Sent from my MI 5s Plus using Tapatalk

 

I keep saying the OEM pre-installed Win 10 is a mess.
Back up the C drive (ssd) with oem Win install, then wipe the SSD, do a fresh Windows install, update it, and do a clean install of latest drivers and chipset from AMD website.

A clean install should take care of all of your issues (if did for me, as I can easily remove all drivers and do a clean driver and chipset install using latest versions and not losing Freesync).

Btw... Freesync not functioning might be a problem with chipset drivers. Have you tried installing the latest chipset separately?
Though it also seems likely the problem could be caused by messed up OEM Windows installation.

 

So as I mention before - finally I successfully install Windows 7 (Ultimate x64). Works like a rocket.
Everything installed perfect except several things:
- Asus Radio Control driver does not exist for Windows 7 (driver from W8x64 don't fit). But I don't use it (don't know why it needed on W7 anyway) and just turn off this in device manager.
- Asus ROG Gaming Center won't install. Maybe W7 version out there but I don't really need it (maybe only for fan boost).
- Problems with a FreeSync - it just don't work.

All devices with drivers. CPU works like it should be. GPU too. Ryzen Master and Afterburner works fine. Everything works like a regular Windows 7.
Really can't figure it out with FreeSync. Method like a manual driver choosing in device manager (Radeon (TM) RX 580) don't work on Windows 7 - it says "this driver does not supported on this system).
Was trying several versions of gpu and chipsets drivers - null. Trying first install several chipset drivers. Trying first install GPU drivers - all for nothing. It's just won't work.
Trying to find a way to hardly manually install the (TM) RX 580 - don't find a solution...
If only AMD roll out official gpu driver for our laptop - would be like a charm... So it's no FreeSync in Windows 7 for now : (
-
For clean Windows 10 install with ASUS product key - my advice is download Windows Creation Tool from Microsoft and just create a USB stick with iso (it will create latest build with all updates).
Worked for me. Key catch up automatically during installation. Then after install (in my personal opinion) do all updates with all devices (Windows will install all drivers). And THEN install latest AMD drivers.
It should do the job. All will working with FreeSync etc. Even better then LTSB version (since it can't catch up all new drivers unfortunately)

 

I haven't had a laptop with window's 10 OEM licence before this one, so far window's 10 has worked flawless for me, I've been using the insider build through upgrading to it. But if I would like to do a fresh install with full wipe of the system how would the licence activation process be? Does it ask? Make a back up through the window 10 user ID? Hard copy the key using the motherboard serial?

Sent from my MI 5s Plus using Tapatalk

 

The product key should activate automatically once you clean-install the same version of Windows 10 that Asus originally put on the laptop (Home x64) because its embedded into the BIOS (or at least, this is the new procedure OEM's decided to go with - hence a distinct lack of Windows product key written on a sticker).
You can follow Terexo instructions on how to make your own up to date Windows 10 clean install.

I had a copy of Windows 10 Pro x64 lying about so I installed that one.

Its possible that Windows Home edition is simply messed up to the point where it causes too much trouble for people on a software level... but I don't think this is necessarily the case.
I think its more likely that Asus broke their Windows 10 Home install with outdated software they tried to jurry rig into the OS and on system restore level outside of Windows.
Remember that while Asus technically has a recovery partition on the GL702ZC, using it is next to impossible as even before the Fall Update, the system simply reported an error at 40% restoration, stopped it, and then threw me back into Windows.

After I force-installed the Fall update, it ended up completely breaking the recovery partition (making it invisible to the restoration program), so it was impossible to restore the laptop to factory defaults since then.

So I figured that I might as well nuke this Windows 10 install and give it a go with a completely fresh Windows 10 installation... and since then, all of my software issues (that I originally experienced with the OEM install) disappeared.

That's why I would advise people to backup their existing OEM install onto a flash drive for example... then format the SSD (plus delete the recovery partitions because they are useless anyway, and you have a backup copy on the flashdrive now), and do a clean Windows 10 Home x64 install.

 

Acer just announced a Ryzen 2700 + Vega 56 variant of their new Helios 500, hopefully Asus will announce theirs soon too.

[Next@Acer 2018] Acer Predator Helios 500, Helios 300 Special Edition gaming laptops announced
Asif Iqbal Shaik, May 27, 2018
https://www.mysmartprice.com/gear/2...laptops-announced-specs-price-next-acer-2018/

"Acer Predator Helios 500, is a completely new high-end gaming laptop.

"There will also be a variant with AMD Ryzen 7 2700 GPU, AMD Radeon RX Vega 56 GPU, and FreeSync display."

AMD RX Vega 56 & Ryzen 7 2700 featured in a future variant of Acer Predator Helios 500 laptop
https://www.reddit.com/r/Amd/comments/8mhvw9/amd_rx_vega_56_ryzen_7_2700_featured_in_a_future/

FURIOUS POWER - Helios 500 - Ryzen 2700x + Vega GPU!!
https://www.acer.com/ac/en/US/content/predator-series-features/predatorhelios500

 

@Deks will you upgrade?

 

I'm pretty happy with GL702ZC and its power/capabilities, so, no.
Besides, I don't have the cash to upgrade even if I wanted to (and it seems really unnecessary because this little Asus beast will last me a while).

Should the GL702ZC break again and Asus ends up giving me a refund (which is unlikely - because they are more likely to offer me a working replacement or an upgrade with an all AMD system) I might though.
Vega 56 in a notebook?
Yes please.
If Acer merely undervolts the Vega 56, they can easily keep its stock clocks and drop the power requirements by quite a bit (around 30% - that should surpass the mobile and desktop 1070 in efficiency and possibly performace). Heck, even if they overclock it properly while undervolted, they should be able to surpass GTX 1080 level of performance slightly while drawing less power than 1080.
Though, running the Vega 56 on stock with an undervolt would likely be better... only overclock if needed, and V56 can already easily game at 2k (so can the RX 580) and 4k.

The cooling in that Helios might be better though... the base is thicker than GL702ZC (however, in all honesty, Asus could have made a FAR superior cooling in GL702ZC without affecting it's thickness in the least - the internal space this thing has is HUGE... cooling shouldn't have been an issue - especially if they decided to move the GPU and CPU further apart and beef up the cooling - and Asus could have unlocked the BIOS too).

It would be nice to have the ability to upgrade the CPU in GL702ZC at least... but Asus said they wouldn't release the needed microcode updates for Ryzen+ (maybe Zen2?)... perhaps someone else might develop a custom BIOS for GL702ZC to include support for Zen 2 on 7nm (now THAT would be great).

Other than that, there's no real need to upgrade the current laptop. The Rzyen 1700 and RX 580 (both undervolted) are pretty powerful in this laptop and can easily run anything on 1080p and High (even maxed out for some games) settings.

 

Why does it take so long for machines like this to happen? Having the option for a 1700X/Vega 56 system when I was after a notebook would have been amazing. AMD need to do what they did with the MSI GX60 and just pay someone to make a high end laptop with their best parts in it so there is always at least one on the market.

 

But the GX60/GX70 was sub-par. CPU bottleneck out the wazoo. AMD drivers woes at the time as well. I knew a lot of people who were unhappy with theirs.

At least AMD has competitive CPU options for notebooks these days.

 

There's probably a negative perception of AMD GPU's due to the skewed power draw on desktops.
OEM's don't take into account the manuf. process AMD used and how limiting it was for GPUs resulting in higher voltages and limited clocks... nor that AMD on average has 40% MORE compute units on Vega for example than Pascal has (which contributes to power draw).

AMD could have also collaborated with OEM's and arranged with them to OPTIMIZE the GPU's on mobile through undervolting, possibly reducing the clocks a bit (due to the manuf. process limitations) and also disabling a certain amount of compute units (which don't affect gaming to be honest - because both Vega 56 and 64 were tested on identical clocks, and both produced exact same performance irrespective of compute units differences) - and computing would still run rather great.

OEM's might have acquired a negative perception of AMD, and Intel bribing OEM's over the years certainly didn't help,

Now, if AMD decided to make their own reference laptops which they optimized themselves... THAT would likely be great to see [although, I honestly cannot be certain if AMD would have optimized the voltages on those GPU's... they might have... but until AMD releases a reference laptop themselves available to the general public, we won't know - thus far we only saw previous reference laptops AMD designed themselves to demonstrate various features on previous generation hardware (before Zen)].

 

And yet, we both know that OEM's usually created laptops with subpar cooling and then put low grade hardware inside, provided inadequate RAM (single-channel usually running on low speed), low resolution screens... and priced them the same as more expensive (and better) equipped Intel solutions.

Dell did the same with Ryzen 2500u and 2700u series (but quick reviews caught them doing this from the very start).
There was also a HIGHLY comprehensive article on how Intel messed up AMD's market permeability and how OEM's ended up pairing AMD with low grade hardware, crippling it wherever possible and overpricing it.

I would have even been interested in the past of getting an A10 for example (as did many other people), but given how I knew that OEM's were crippling AMD hardware...

 

It doesn't help that AMD also cheaped out on power monitoring circuitry, which famously led to the RX 480 launch fiasco:

I find this doubtful. Do you have a source?

 

That didn't really apply to the GX60/70. It was the same chassis and cooling as the Intel+Nvidia GT60/70 and priced more competitively for the specs.

The issue was the A10 was just that bad and crippled AMD's flagship mobile GPU at the time.

https://www.anandtech.com/show/7111/amds-a105750m-review-part-2-the-msi-gx60-gaming-notebook/6

 

AMD did officially start recognizing the issue and since then amended their reports of power draw.
Also, AMD is a much smaller company with fewer resources... it might be one of the reasons they did this with power monitoring circuitry.
Besides, Nvidia has a tendency to draw much more power beyond their stated TDP as well when stressed, and certain things aren't being reported either.

At any rate, undervoltng AMD GPU's works, and it cuts power consumption by quite a lot. We also know that a slight underclock can also help bring down power consumption further as the manuf. process (14nmLPP) is not Polaris or Vega's friend in regards to clocks and power draw (however, it is also worth noting that AMD managed to achieve SAME performance with lower clocks than Nvidia does, indicating higher IPC, but at said clocks, the manuf. process works against them, which resulted in lower yields, producing higher power draw in the process - and also the fact AMD introduced higher number of compute units on their GPU's).

Here you go:


Also, I should have been more precise in saying that equalizing for clocks did not affect GAMING performance between V56 and 64...
Obviously in compute tasks, there would be a performance difference in V64 favor...

 

I know that the A10 wasn't as powerful as the Intel cpu portion was, but even crippling the 7970m, it provided a relatively good performance.
MSI still priced the G60 too much, and the fact that they included only a single RAM stick (which is how the laptop was tested) and had no SSD in it didn't help the performance numbers.

AMD APU's (both CPU and iGP) do depend quite a bit on whether RAM is running in single or dual-channel mode... along with RAM speeds and latencies (low latency and high speed RAM worked best).
Here's something else to look into for A10 4600M (the same trinity based APU used in G60) :
http://forum.notebookreview.com/thr...ested-with-various-ram-configurations.712179/

As you can see, running dual channel does affect performance on APU's, usually on their iGP, but it also affects CPU performance somewhat, and we don't know how much of an impact there might have been on dGPU as a result of A10 having access to dual-channel and more bandwidth.
Most of the Trinity and other APU configurations were tested with single-channel RAM which affected performance for the worse in games... this could have also translated to CPU related tasks given how dependent those apus were on RAM and bandwidth they provided.

 

Can you elaborate? One of the points Buildzoid made in the video was that Nvidia does have accurate power monitoring compared to AMD. The GPU Power sensor in HWiNFO (read through NvML API) is a pretty accurate total board power for Nvidia and is used for GPU Boost 3.0. Instantaneous spikes above the power limit are possible in power viruses like FurMark, but for better or worse, Pascal sticks very tightly to its set power limit.

Well no, the GX60/70 was less expensive than similarly configured Intel+Nvidia GT60/70 with a better CPU and worse GPU, as mentioned in the review. Also, the other GX60 with A10-4600M had dual channel RAM and SSD and it didn't really affect gaming performance: https://www.anandtech.com/show/7111/amds-a105750m-review-part-2-the-msi-gx60-gaming-notebook/4

 

i am using it with Fedora. i managed to get the recent kernels compatible with our touchpad.. other than that, no big issues. Very happy with it

Sent from my MHA-L29 using Tapatalk

 

Have you managed to get the touchpad working on Win7 yet? I haven't put too much effort into it as I've been busy and just using a mouse, but it seems like it's I2C connection and we need a driver for it.

 

Forgot about touchpad. Unfortunately I don't find a solution for working touchpad. But I don't really searching good because I use only mouse or (very rare) I just use keyboard. Have you tried to search drivers by device ID from device manager or just install drivers by Asus (they might fit Win7 as well)?
-
Also need mention about GPU work on Windows 7. So I don't find a way to get FreeSync to work (even find a way to hardly install (TM) RX580 with BSOD). BUT another thing is that on Win7 GPU is loud. I mean in SOME applications (standart Chess or old games like Max Payne) the drossels (I think) make a noticeable squeaking sound. Really don't know why and how to fix it. For now I'm on Windows 10 LTSB 2016 and everything just fine. Maybe later I will work around again with Windows 7 (still it's fully working, with no difference in performance and even better in some situations).

 

Yeah I spent a couple hours looking at it yesterday - no luck for it as far as I can tell. I got everything (including the wireless radio control and ROG center etc) working apart from FreeSync and the touchpad. Windows 7 seems to lack I2C support and the driver we need is ELAN 1203.

Like you I ended up reverting to LTSB. I've been playing with DXVK on Linux recently too, and it looks like I may be able to dump Windows for gaming now

 

I just rolled back my update as well. Same issues.