Asus ROG GL702ZC owners lounge

Btw people, newest Radeon Adrenalin drivers:
https://www.guru3d.com/files-details/amd-radeon-adrenalin-edition-18-3-1-driver-download.html

Just use the update function inside the drivers and use Express install option.
Some nice updates for RX 580 as well.

 

I would have to say that should be unusual, but with all the reports here showing the same 17-18% wear level, it appears to at least be common. I am checking with ASUS about this...will report their answer.

 

Great, tnx.
Also, if you speak on behalf of HIDevolution, perhaps you could also inquire about Asus releasing BIOS updates for Ryzen+, Ryzen 2 and Ryzen 3.

 

ASUS has not informed us regarding BIOS update roadmaps...nor do they ever do so. We just have to wait to see when they are released.

 

When did you purchased yours? I bought mine on late January and my GPU's ASIC quality is 83% and never goes beyond 84C even on the most extreme cases nor never throttles. My CPU never throttles but can reach 94C with Prime95 but using games and other stuff, rarely goes beyond 60C.

 

Wow that's terrible.

Aside, @Deks you mentioned undervolting. Are there options for that in the BIOS, or did you use software in Windows? If it's the latter, are the changes dependent on the software and thus Windows? I'm especially interested in undervolting that is persistent and applies in linux.

 

I purchased mine in October 2017... so mine is practically the first unit available on the UK market... but the GL702ZC was developed before that I would imagine... just waited a while for UK vendors to pick them up I suppose, which is probably why it sat in a warehouse before being sold.

Anyway, my temps are similar, if not slightly higher on stock, but otherwise, with an undervolt and overclock on the CPU (3.3GhZ and 1.05V - though I can undervolt the 3.2GhZ to 0.98125V), I can get about 84 degrees C maxed out on the CPU like that... and about 75 degrees C on GPU with a -93mV undervolt on the core (when maxed out of course).

I know... I was surprised by hoe much the battery degraded vs the time I got the laptop.
If this continues, the battery will likely stop functioning before or just after the first year of owning the laptop (lucky for me I got a 2 year warranty though).

BIOS on GL702ZC is locked (shame on you Asus)... I used Ryzen Master Threadripper to undervolt and overclock the CPU (3.3Ghz on 1.05V), and latest MSI Afterburner to undervolt the GPU (-93mV on the core is stable for me across all workloads).

 

The wear level is measuring actual potential against "documented" potential, what the spec says. This isn't always correct, variances in manufacturing and vendors used to supply the battery may not meet spec as programmed.

The out of the box reading seems to be the same for everyone, so those batteries were built about 15% under programmed spec. Typically I see 2%-8% wear level stabilize after a few weeks, for batteries closer to programmed spec.

Also if you've not been running the battery down far enough, and charging back to full, the battery is unconditioned and may be better / worse than out of the box after a few charging cycles.

Some weird things have been reported on Asus batteries lately, with the wear level showing continual drops on some GX501's like @mason2smart 's unit that he is trying to get a new battery replacement, so watch for delta changes over time, rather than worry about initial numbers.

I'd do a couple of full (15%-30%?) discharges manually, or if Asus has a battery conditioner that supports your model run that, and see where the wear level ends up, then watch for precipitous drops over time from there.

 

I just checked my battery wear level, and HW monitor (latest beta) says it's at 16.4%.
Huh... not bad when you consider my laptop is now almost 5 months old... but they DID replace the motherboard (although I don't know if Asus also replaced the battery)... but it's also discouraging, because the rate of degradation is not exactly encouraging.

I guess I'm one of the luckier ones with a 2 year warranty, so, I can technically request a replacement unit just before the warranty expires if the degradation continues to progress at the same rate every 5 months or so... and 'complain' about the speakers vibrating too.

Dunno... we'll see.
Technically speaking, while the battery is not replaceable... its still connected to the motherboard with a cable which can be safely detached/reattached if I'm not mistaken... so we should be able to do it ourselves as well (if something happens to the battery outside the warranty).
But if you're within the warranty... then its better to ask Asus to fix that... or just send a replacement unit (which would be the fastest option, because repairs can take time).

 

@Deks How does the undervolting affect idle temps and noise/fan levels?

Are the settings made by Ryzen Master and MSI Afterburner dependent on those software being present? i.e. are they persistent? It would be great if you could test if the changes remain in effect under linux.

 

Warranty details for batteries should be checked. Usually that's considered a consumable and excluded from the regular warranty. Batteries usually have a lower coverage period, often 90 days to 1 year max.

 

Undervolting affects load temps only. Idle temps remain same.
Both ryzen master and MSI afterburner need to be active for undervolting to stick. For persistent voltages, you'd need to affect them on a hardware level, and right now, we can't do that if I'm not mistaken.

I don't have Linux I'm afraid.

Perhaps, but I don't think this applies to integrated batteries as they are essentially part of the unit itself and should be covered under warranty regardless.

 

Best to read your warranty.

 

Oh... well if the software needs to be active, that answers the question.

Is there any 3rd party or custom/unlocked BIOS?

 

Not as of yet, but software undervolting is not a problem as its not heavy on the memory or system resources at all (practically non existent).

And considering this is a laptop with a locked BIOS, I think doing this much (both undervolting the cpu and gpu and overclocking the cpu) is pretty good for the time being (laptops obviously don't have same treatment as desktops).

Custom BIOS might come later from someone else, but right now I'd settle for Asus giving us new BIOSes for ryzen+, 2 And 3

 

Though, it may be possible to affect hardware undervolting on the GPU through vbios modifications... But I hadn't done that with the RX 580. Only on my old nvidia 9600m gt.
Software undervolting is a bit better because you can avoid bad bios flashing which can otherwise turn your laptop into an expensive paperweight if some of the settings are wrong or something else goes wrong during flashing.

 

It's more of a problem for linux users who virtually never boot into Windows..

In my experience, BIOS updates come at the beginning of the system's lifecycle to fix bugs after launch... and after that don't see much if any attention from manufacturers.

The biggest irritation on my previous Asus was that the BIOS did not support the full RAM capacity of the hardware. Despite my complaints, they never fixed that.

 

Actually, someone here tested GPU undervolting under Linux for GL702ZC.
I suggest you look near the beginning of this thread for clues on the user who did it (or was it in another thread for GL702ZC?).

Don't know about CPU undervolting under Linux though.

Yes, its a common practice for OEM's to initially support a laptop and release BIOS updates early in the products lifecycle, but then decide to abandon it afterwards.
However, Asus has a slightly better track record for keeping up updates for some of their laptops.
I guess we will need to wait and see.

 

I don't recall reading about it in this thread.

Incidentally, where in its lifecycle is the GL702ZC model? Relatively few stores have it to begin with, stock seems limited, and some stores have even listed it as discontinued.

 

@Deks
Above you mentioned that the battery can be unplugged from the motherboard, and while that's true I think the main reason is that you can discharge any remaining juice from the motherboard by pressing the power button after the battery has been unplugged. Not doing so may result in an unhappy experience that I encountered with my old m15x where thermal repasting the CPU resulted in a blown condenser that made the fan's stop spinning, 100 Euro repair cost.

Sent from my MI 5s Plus using Tapatalk

 

Well, I just could not resist, so I opened her up late last night, installed the 960, cloned the hdd, swapped the hdd out for the 850, removed the stock ram and installed 2x16 GSKill 2400's I had purchased months ago during a "NuevoHuevo" flash sale and that were just hanging about unopened, pulled a 1421 R15 score, installed Ryzen Master,...then noticed what time it was and retired for the evening.

Unsure how I survived this long without an NVMe...boot times are ridiculously fast, and apps load in an instant. I'll perform some measurements/benches later today.

 

Yeah, how could you survive with an Hdd ?

 

Don't misunderstand me...I've been using SSDs for years, but since I was going to install my own 960 I opted not to get any SSD, and since the unit had to be configured with the SSD or an HDD I opted to order it with just the stock HDD - and so the unit came with Win10 installed on the HDD.

What I had not yet experienced was an NVME that clocks in at about 3200/1900...I was used to 550/500...and man oh man, does it ever make a difference!!

 

I already know this You as others have hardware setups in sig

 

I'm surprised, is there really much difference between NVME & SATA 3 SSD? I thought I read in lots of places that boot times, application opening times are pretty much the same between the two, and NVME is mostly useful for video editing & large file manipulation?

 

Well, compared to my P570WM with SATA SSDs in raid 0 (see signature - thanks @Papusan ), this rig flies. Just stating my perceptions based upon my experience...your mileage may vary...

 

Yeah, that's right - you gotta compare it in the same rig. I'm fairly certain that boot times & application load times are the same between NVME and SATA 3 SSD's, but manipulating large files then the NVME will win significantly - most people don't do that though or not very often, and if you're one of them then you probably know. (And RAID 0 is only good for large file manipulation too, not for boot times or application load times - in fact RAID 0 can decrease performance of random 4K reads, which can hurt boot times & application load times slightly, if I recall from reading reviews).

 

How full is the SATA3 SSD? Full vs empty changes performance as well. Of course testing on a different system changes the picture too.

From what I've read, NVMe will still be faster in boot times, but not by much.

 

I just recently purchased my GL702ZC and I've been able to overclock the CPU quite well with Ryzen Master but the RX 580 GPU is hard-locked to 1077MHz regardless of overclocking software. I've read all the warnings about heat and power consumption but I'd still like the ability to overclock the GPU, if only a little. The fun is in being able to tinker!

So it looks like I'd have to flash the VBIOS to make the GPU overclockable, and there is a mention on page 19 of a user stating that he was able to modify and flash his VBIOS for 1250MHz, so I feel this should be possible.

I've used Polaris Bios Editor to edit my stock VBIOS, changing only the "Powerplay" values so far. I've changed the Max GPU Frequency from 0 to 1600 and the Max Memory Frequency from 0 to 2300. If I understand correctly, flashing this modified VBIOS would not change the chip's default behavior in any way, but would make it possible to overclock using software. Please confirm or correct me if I'm wrong.

I'd really like to try this while I'm still within my exchange period, just in case flashing bricks the PC, but I'd also like to hear the community's thoughts and suggestions first. After all, the last AMD GPU I modified VBIOS on was the 7970M so I'm feeling a little rusty here.

 

I honestly wouldn't know, but I'd be interested to see if it works if you want to 'experiment' and see how the flashing holds up.
Plus, it might be possible to have Radeon Wattman then appear in drivers as well if you are successful.

Polaris BIOS editor does seem like the best bet to do it... but I'm not sure how you'd go about flashing the VBIOS then back onto the GPU.
Are you also using Windows 10 or Linux?

 

I'm running the stock Windows 10 OS. Actually I installed the RX580 reference driver from AMD, version 18.2.3, and Wattman actually does show up but I still cannot change any values. Right now I'm thinking of putting ATIFlash 2.77 on a FreeDOS bootable flash drive and using that to flash the modified VBIOS I've saved. I know there's a GUI version of the same utility but flashing VBIOS within Windows makes me more nervous than doing it via command line.

 

How in the world did you get Wattman to appear?
As for changing values in it... I think you need to enable manual voltage control for both the core and VRAM.

 

Hello guys... First, sorry for my english. I'm almost buying this thing, however I have a few questions for you:

1) why gc extreme on the GPU and conductonaut on the CPU? Why not conductonaut on both?

2) anyone tried the 960 pro nvme?

3) which ram model do you recommend to put on the empty slot to enable dual channel at full speed?

 

1) No IHS on a GPU = too risky. Listen to what this guy has to say:

2) I haven't, but I'm very happy with my 960 non-pro thus far.
3) The module that came in my laptop is a Samsung M471A2K43CB1-CRC - so, if I hadn't swapped it for a set of GSKill 2400 that I had purchased for a rainy day, I would have just gotten another one.

 

1.) As Zdroj already mentioned, it would likely be a bit too risky to have it on the GPU... besides, GC extreme would be more than enough.

2.) I hadn't, but Zdroj already said he has a 960 non-pro ssd in his system, so it should be giving back pretty good performance as far as moving large files goes (and yes, I'm merely repeating what was already mentioned).

3.) Depends on which RAM you have inside. I would imagine that Asus shipped the unit with 2400 MhZ RAM, so, any RAM stick of same size and frequency should be fine. You can check CPU-Z for RAM frequency, timings, and amount. to help you choose appropriate stick.

 

To be honest, I wasn't actually trying to get Wattman to appear at the time. Rather I was trying to get the system to behave better with the LCD overclocked to 90hz, but I'll try to recreate the steps as I remember:

1. Download the reference display driver for the DESKTOP RX580 GPU, version 18.2.3, directly from AMD.
2. Open Windows Device Manager and delete the Radeon Display Adapter entry. Check the box to also remove the associated driver software.
3. Install the newly downloaded display driver.
4. Reboot system.
5. Open Device Manager back up, open the new Display Adapter entry, click Update Driver, choose "Browse my computer...", choose "Let me pick...", choose "Have Disk" and then enter in the location where your driver files were unzipped to, while also unchecking "Show compatible hardware." In my case the driver location was: C:\AMD\Win10-64Bit-Radeon-Software-Adrenalin-Edition-18.2.3-Feb22\Packages\Drivers\Display\WT6A_INF
6 The GPU I chose is called "Radeon(TM) RX 580." NOTE: There are TWO spaces between (TM) and 580. There are many very similar options so you probably have to pick the exact one. For instance I've read that choosing one without the (TM) disables FreeSync.
7. Accept warning(s) and rebooted again.
8. Download and install MSI Afterburner if you don't already have it.
9. Open Afterburner's settings and check "Extend official overclocking limits" in the "AMD compatibility properties" section.
10. Reboot again. Cross your fingers and hopefully you'll have Wattman after the reboot.

Note that I'm not able to actually change anything in Wattman but it's there and it looks cool!

UPDATE: I left out some of the steps when first posted; fixed now.

 

@Deks @zdroj Thanks for the info.

Now, I have one last question. I watched some teardowns and I think I saw thermal paste on the GPU memory and everywhere else, that is, no thermal pads.

Is it true? If it is, I think it is a bad idea changing that.

 

Yes I believe it's true. I haven't pulled off my heatsink yet - I will do this when I can afford a RAM upgrade, as the heat pipes block the RAM slots. Still I can see areas where the paste is visible around the edges so I feel comfortable confirming that it's all paste.

Before you follow my steps from the last post, I updated the post because repeating the same steps on my PC removed Wattman from Radeon Settings. If I get it back I'll post how I did it, but I was never able to change anything within it anyway. I think it'll take a VBIOS update to enable that.

 

I got it!! Enabling "Extend official overclocking limits" in the "AMD compatibility properties" section of MSI Afterburner's settings brings back Wattman after reboot. Still cannot change anything (it reverts back on clicking Apply) but it looks cool that it's there!

I wonder if applying this setting on the normal Asus display driver would have the same effect. I don't think it did but I cannot remember for certain whether I'd tried before updating the driver.

 

Hmmm...not sure. I had some thermal upgrades performed by HIDevolution when I purchased my rig, so I really can't offer you much here...but maybe @Donald@HIDevolution or @thattechgirl_viv can.

Say...speaking of her...I wonder what ever happened with the suggestions of @hmscott? The silence is deafening!! I suspect she just doesn't want to part with the gratis "super rig" I suggested she build for @hmscott if those suggestions pan out...!??

Also, I'm not sure if @Donald@HIDevolution ever got a response from Asus re: the battery concerns, but after I let mine run down to about 67% I began recharging it, and it has registered "Fully charged (100%)" for the past 2 days...go figure ?!

 

Thanks @AngieAndretti for the info (and for the Wattman).

About the thermal paste/thermal pads issue, then I would not change that.

My currently laptop is a MSI GX60 1AC (FS1 socket on CPU and MXM). It has thermal pads on the memory of GPU (HD 7970m).

After seeing the teardown of the GL702ZC (and seeing thermal paste on the memory modules), I made a little (and very silly) experiment: changed the thermal pads to thermal paste.

I do not turned on the machine of course - just reassembled and disassembled. The result was exactly what I expected: The thermal paste on the memory modules was not spread. The problem is that as the heatsink was not made for thermal paste, and it has 1 millimeter more in height on the GPU than in the modules to perfectly accommodate the thermal pads.

In conclusion: Maybe doing that in reverse (changing the thermal paste to thermal pads), would cause the heatsink to do not fully contact on the GPU module (as it gets more height due the thermal pads on the memory).

BUT - To be fair, if the guys at HIDevolution did that, then I think it is safe and I am wrong

 

ASUS has been doing this for years...it is a mess. If anyone orders our Thermal Materials upgrade, we replace the paste with good thermal pads, unless they order Fujipoly Extreme thermal pads.

 

I will let @thattechgirl_viv answer once she has the results. However her offer was lunch and dinner

The wear level issue seems to be present with every one we take out of the box. It has been reported to ASUS...we will just have to see if they come back with a proper response.

The charging level is a different issue. However remember I said it will stop charging anywhere in the 90's to 100%. "...anywhere ..." means that the charging level isn't necessarily consistent. It can vary from charge to charge.

 

Which pad thickness should we be aiming for?
0.5mm or 1mm?
I have the 0.5 mm pads, so I'm wondering if they would be a good fit for the GPU RAM and VRM's?

 

You could try right clicking on the app icon and select "Run As Administrator", see if that increases permissions enough to make a change.

 

The best way to tell is to apply them, and see it they fit with proper compression, or you need thicker ones. A lot will depend on exactly what pads you have.

 

Right, but which ones have you used for the GPU VRAM and VRM's?

 

You know that is a question for @thattechgirl_viv, our Production Manager.

 

Hi all,

I want to share my expirience and I thought I could get here a shot from you guys, so let start:

I bought this laptop about a month ago and I am really impressed. The SSHD was a little bit anoying, so I
put in a 500GB 850 EVo and a 960 250GB M2.

I didnt like the temps and because I am a small enthusiast, I started to undervolte the GPU and CPU to following stats:

GPU -0.93mv
CPU 3.6 [email protected] all cores

For the GPU I used MSI Afterburn what works great to undervolt this laptop, also I managed it to set the gpu clock higher, but after some seconds I got a crash. Maybe need to flash the gpu to increase the clock, but for now I will keep it like it is.

For the CPU I just used the Asus Zen States and it still working like a charm. Btw it is a R5 1600.

Although the temps were better, I was going on and started to disassemble that thing and put liquid metal on the CPU and GPU (Grizzly conductonaut). I used nail polish to prevent a circuit shot in case the liquid metal get it touch of something else than the GPU/CPU.

https://ibb.co/eHua07

After a fresh start the idle temps need about 20 minutes to reach 45 °C where the fans start to turn on. Directly after starting windows I get now about 25°C for CPU/GPU. When I used the automatic mode in ROG Gaming Center the GPU hit about 77°C in Superposition and the CPU about 94.5°C in prime95. Thats why I really dont like the fan curve what asus has set up.
And here I get some problems:

I found the actually really nice program called NBFC (Notebook Fan Control) which read and write into the embedded controller. First I tried to find the E/C on the motherboard and I thought I find it, but no where in the world wide web I can find these controller:

https://ibb.co/kaFOL7

"ITE8171FN"

I am also not really sure if that is the ec, but it is a little bit strange that I cannot find anything about it.
So I started to use the RWEverything program and I also found the values for RPM, temp CPU/GPU and fan stages.

Here I am stucking right now and couldnt get it managed to control the fans,I dont know why it is not working but maybe someone of you tried already to control the fan speed and make his own fan curve. IMHO the ROG Gaming center is completly crap if you want to set up your own fan curve and If I read it right, some Asus laptops are indeed able to set up your own fan curve.

 

Did you by any chance follow my advice in using MSI to undervolt?
If so, then you might want to know that my RX 580 is also stable at -93mV.
Depending on your ASIC quality, your voltage values may fluctuate. My ASIC is only at 71.7% (hence why its only stable at -93mV)... and if your ASIC quality is higher, you might also be able to undervolt better than -93mV (use mine as a basic guideline as its more likely you would have better silicon in general if your CPU is any indication).

My 1700 crashes in Cinebench at 3.6GhZ and 1.16875 V - it needs a notch higher (1.175 V) to complete successfully. Though, I'm unsure as to the long term stability on that voltage... because Cinebench is one thing, and Primark 95 something else.
Though I can probably run 3.5GhZ on 1.168V or 1.1625V.

As for the fan curve... NBFC is a confusing program to say the least because its not easily understood, and searching for viable values on GL702ZC is like finding a needle in a haystack.
Plus, the GL702ZC has a locked BIOS... it's probably why we're having a hard time affecting fan curve values in the first place. - though, undervolting both the CPU and GPU helps a lot in managing the fans and noise and makes the unit far more viable without headphones.

I was wondering what your temperatures are with Conductonaut running on the CPU and GPU when both are stressed to the max?