GL702ZC Gpu Seriously Underperforming

[yemred]

Hello Rog family,

I recently bought a doa gl702 (motherboard was changed) and I've been having issues performance wise since then. I am a newly graduated architect and I chose this laptop because of the cpu power and relatively good gpu power. Anyway in the first three days after purchase I got black screens every now and then and one time keyboard completely stopped working for some reason and started working again after clear reinstallation of all drivers. Then I tested gaming performance with Dishonored 2 and I felt like it was getting sudden lags like one in every 5 secs. I was getting a fps range of 35-50. Then I started rendering with Lumion (a gpu intensive rendering program for those who are not familiar) and I got 10 fps which is a very low value comparing to other Lumion users with exact same laptop/specs who gets around 20 fps. I did a gpu-z logging while rendering and I think there is a thermal throttling going on. You can see it here: https://gifyu.com/image/Tcnj As you can see when it hits the 80C it lowers the gpu core clock and makes a fluctuation as well as sudden drops in memory clock.

One other thing I did was a 3dmark TimeSpy test and it gave me a significantly bad result if not the worst. https://www.3dmark.com/search#/?mode=advanced&url=/proxycon/ajax/search2/cpugpu/spy/P/2220/1154/5000... Here you can see all the results of the GL702 model and my score is two rows high from the bottom one out of 140 results. Here is the comparison link https://www.3dmark.com/compare/spy/5022976/spy/4080375/spy/4373009/spy/4176899/spy/5020361/spy/38607... and you can see there is a result of exact same specs including memory which got 3604 while I only got 2184.

Long story short, do you think something is wrong with the laptop and I should take the laptop in for service? I still have 23 months warranty but I am not sure they will think these issues are worth considering. The gpu is coil whining on top of that.

[deksman2]

Sounds like a hardware problem to me.
Though, the service centre might not have applied the thermal paste properly after they repaired the laptop the first time around.
I'd suggest RMA-ing it again.
Detail the problems with the GPU and see if you can find any other problems that might have been happening which you can mention to Asus in the online RMA form.

My GL702ZC broke down the second time and I had to mention to Asus that the speakers were producing vibrations (so they were not functioning properly), GPU produced minor issues during browsing (such as green lines which appeared momentarily but disappeared once you scrolled down), and of course the primary problem was the fans started running at 100% again and refused to cycle down (and 45 mins after that happened, the laptop died completely).

I asked them to have a look at the cooling and see if they can replace the cooling assembly, but I think they might have only replaced the fans (not the heatpipe system).

Asus I think made a VERY bad design choice with the cooling... that or they had a bad batch of motherboards which end up failing after a while.
I'm not satisfied that this happened twice already.
Also, the replaced motherboard failed again while they were testing the unit, so they had to get ANOTHER motherboard.
Where are they getting the motherboards that keep failing? The quality control is atrocious.

And of course, Asus refuses to give me a replacement or a refund unless the unit undergoes 3 or 4 failed repairs... which is stupid.
I voiced a lot of engineering concerns about the unit, but all of them fell on deaf ears.

Anyway, what you can also try is to undervolt your GPU using MSI afterburner (the stable undervolt I can get on the core is -93mV).
Also, try to upgrade your motherboard BIOS if you hadn't already.

But ultimately, if your GL702ZC cannot perform adequately WITHOUT an undervolt, its likely that the service centre messed things up in terms of the thermal paste.

RMA the unit and tell them its throttling, the GPU is faulty and not producing the performance it should (this could prompt them to replace the motherboard)

[yemred]

Hey deksman2, thanks for the reply. I'm sorry to hear you had bad luck with the product as well. I too think that Asus didn't make enough quality control for this model and many users complain about hardware issues which makes me think what will happen after a couple of years using the product. I would like to ask you, do you think the gpu-z logging shows a thermal throttling? I mean thermal throttling at 80 °C doesn't make sense to me, shouldn't it be like 90 °C? And regarding your suggestion, how much could undervolting improve performance? As you see timespy results are pretty bad for my device and I was wondering if undervolting could make significant impact on the result.

[yemred]

deksman2 wrote:
Sounds like a hardware problem to me.
Though, the service centre might not have applied the thermal paste properly after they repaired the laptop the first time around.
I'd suggest RMA-ing it again.
Detail the problems with the GPU and see if you can find any other problems that might have been happening which you can mention to Asus in the online RMA form.

My GL702ZC broke down the second time and I had to mention to Asus that the speakers were producing vibrations (so they were not functioning properly), GPU produced minor issues during browsing (such as green lines which appeared momentarily but disappeared once you scrolled down), and of course the primary problem was the fans started running at 100% again and refused to cycle down (and 45 mins after that happened, the laptop died completely).

I asked them to have a look at the cooling and see if they can replace the cooling assembly, but I think they might have only replaced the fans (not the heatpipe system).

Asus I think made a VERY bad design choice with the cooling... that or they had a bad batch of motherboards which end up failing after a while.
I'm not satisfied that this happened twice already.
Also, the replaced motherboard failed again while they were testing the unit, so they had to get ANOTHER motherboard.
Where are they getting the motherboards that keep failing? The quality control is atrocious.

And of course, Asus refuses to give me a replacement or a refund unless the unit undergoes 3 or 4 failed repairs... which is stupid.
I voiced a lot of engineering concerns about the unit, but all of them fell on deaf ears.

Anyway, what you can also try is to undervolt your GPU using MSI afterburner (the stable undervolt I can get on the core is -93mV).
Also, try to upgrade your motherboard BIOS if you hadn't already.

But ultimately, if your GL702ZC cannot perform adequately WITHOUT an undervolt, its likely that the service centre messed things up in terms of the thermal paste.

RMA the unit and tell them its throttling, the GPU is faulty and not producing the performance it should (this could prompt them to replace the motherboard)

Hey deksman2, thanks for the reply. I'm sorry to hear you had bad luck with the product as well. I too think that Asus didn't make enough quality control for this model and many users complain about hardware issues which makes me think what will happen after a couple of years using the product. I would like to ask you, do you think the gpu-z logging shows a thermal throttling? I mean thermal throttling at 80 °C doesn't make sense to me, shouldn't it be like 90 °C? And regarding your suggestion, how much could undervolting improve performance? As you see timespy results are pretty bad for my device and I was wondering if undervolting could make significant impact on the result.

[deksman2]

yemred wrote:
Hey deksman2, thanks for the reply. I'm sorry to hear you had bad luck with the product as well. I too think that Asus didn't make enough quality control for this model and many users complain about hardware issues which makes me think what will happen after a couple of years using the product. I would like to ask you, do you think the gpu-z logging shows a thermal throttling? I mean thermal throttling at 80 °C doesn't make sense to me, shouldn't it be like 90 °C? And regarding your suggestion, how much could undervolting improve performance? As you see timespy results are pretty bad for my device and I was wondering if undervolting could make significant impact on the result.

The GPU should not be thermal throttling at 80*C... not even 90... but the thing is that GL702ZC should not even allow the GPU to reach 90 degrees C (if the GPU was maxed out, the fans would run at maximum and likely bring down the temperature of the GPU to about 64*C).
So if it does throttle, the fans might not be cooling it adequately, or as I said, the thermal paste was not applied properly.
RMA it.

As for how much undervolting improves performance.
Difficult to say.
I undervolted the GPU to drop load temperatures (which can drop between 5 and 10 degrees C)... it helps in reducing fan noise during gaming.

You can also undervolt the CPU using RyzenMaster.
I'm using the Creator tab in RyzenMaster, and have set the CPU to 3.2GhZ across all cores and voltage to 0.98125V (this is stable voltage for me at those frequencies).

As I said, upgrade your BIOS if you hadn't already (latest version is 305) and try undervolting... and your best bet on improving performance would be to have 2 RAM sticks running in dual-channel mode (this is the single greatest performance boost you will see in games as Ryzen is dependent on dual channel RAM).

But, if you see same results after bios upgrade, just RMA the unit.

[yemred]

deksman2 wrote:
The GPU should not be thermal throttling at 80*C... not even 90... but the thing is that GL702ZC should not even allow the GPU to reach 90 degrees C (if the GPU was maxed out, the fans would run at maximum and likely bring down the temperature of the GPU to about 64*C).
So if it does throttle, the fans might not be cooling it adequately, or as I said, the thermal paste was not applied properly.
RMA it.

As for how much undervolting improves performance.
Difficult to say.
I undervolted the GPU to drop load temperatures (which can drop between 5 and 10 degrees C)... it helps in reducing fan noise during gaming.

You can also undervolt the CPU using RyzenMaster.
I'm using the Creator tab in RyzenMaster, and have set the CPU to 3.2GhZ across all cores and voltage to 0.98125V (this is stable voltage for me at those frequencies).

As I said, upgrade your BIOS if you hadn't already (latest version is 305) and try undervolting... and your best bet on improving performance would be to have 2 RAM sticks running in dual-channel mode (this is the single greatest performance boost you will see in games as Ryzen is dependent on dual channel RAM).

But, if you see same results after bios upgrade, just RMA the unit.

Thank you for the detailed response, it seems you are the only one trying to help, so much appreciated. One last question, could you share your afterburner values please? I am only changing core voltage to -93 mV but the software allows me to increase core clock value as well, Im guessing if I increase the value it will still work on 1077 mhz max, right?

[deksman2]

yemred wrote:
Thank you for the detailed response, it seems you are the only one trying to help, so much appreciated. One last question, could you share your afterburner values please? I am only changing core voltage to -93 mV but the software allows me to increase core clock value as well, Im guessing if I increase the value it will still work on 1077 mhz max, right?

My afterburner does not allow me to increase the core clock... if I try and press 'apply'... it just resets itself to 1077MhZ.
My voltage is set to -93mV on the core... that's stable and works for my gpu (just apply that voltage on your GPU and hit ''apply'')
I suggest to keep the voltage on that level (it really depends on your GPU if you can drop more voltage... but if you can modify your clocks, then try doing that while undervolted.