Fluctuating GPU frequencies problem

My GPU is apparently garbage and it cannot reach its full potential for an unknown reason due to clocks fluctuating wildly (though in all honesty, I have no idea if this is hardware or software related).

Running Furmark GPU stress test for a limited period of time, I've noticed that the GPU cannot sustain maximum clocks at all.
Its clocks wildly fluctuate from minimum to medium to maximum (you can see screenshot of running Furmark benchmark how the GPU frequency graph is fluctuating and an example of how low the frequencies are).
This is NOT normal behaviour and it explains why Darksiders III was performing poorly too.

I don't understand how/why is this happening.
Even before the RMA, the GPU was displaying similar behaviour in Furmark (fluctuating clocks, but they were fluctuating from middle to high as opposed to minimum to high) but only after it went past certain temperature levels (above 73 degrees C).
And as you can see, the GPU is nowhere near what I would call 'dangerous' temperature levels and after the RMA, this GPU clock fluctuation is happening since the GPU is stressed).

Something seems to be preventing the GPU from leaving its minimum state and maintaining its boost clocks... I tried other drivers (with same results).
I tried turning off Power efficiency and Chill off ... same thing is happening.
I tried even increasing power limit to 50% with no difference.
I tried undervolting, which lowered the temperatures to around 73-74 degrees C but clocks were still fluctuating like mad (so, no difference in actual performance).

The GPU does not appear to be throttling... or is it?
I don't know.
Could it be a Windows issue?
I hadn't tried reinstalling the OS... so that might also be contributing to or causing the problem (but as you may know, reinstalling the OS is a last resort as its a time consuming process - I can do it of course, but I need help in trying to narrow down what's causing the problem).

EDIT: Someone on AMD forums had a similar problem with their desktop GPU and the only way they solved the problem was to open the GPU and replace the thermal compound.
They said the GPU was overheating and that the displayed temperature was wrong.
My issues begin few seconds after running FurMark GPU stress test.

 

AMD throttles its clocks to bare minimum in Furmark test to avoid destroying its GPU. I thought AMD users knew that. I had 5750 radeon and it had same issue as well. Only in Games and rendering you can max out OC values. Try Luxmark and use max OC and +20-40% power limit, your score should have improved vastly.

 

I wasn't running any OC.
Darksiders III game exhibits identical behaviour as Furmark.
Also, I ran Superposition behchmark and according to internal measuring of clock behaviour, the GPU core clocks fluctuate like mad (same as in Furmark and Darksiders III).

 

Star Trek: Online seems to work fine (but its also an older game that's not as demanding).
When I also ran Luxmark, I opened MSI afterburner and noticed the GPU clocks holding steady at 1077 MhZ.

So, why are the GPU clocks fluctuating on Darksiders III the same as they do in Superposition and Furmark benchmarks?

Are there certain things in these benchmarks and game that my GPU cannot handle?
That's doubtful, because desktop users of RX 480 and 580 run Darksiders III without the same problem... so my GPU should be able to run the freaking things too without problem either (yes at slightly reduced performance, but it should be able to handle it regardless).

 

Here's an uploaded screenshot of my Luxmark test.
As I said, the GPU clocks can maintain their maximum boost no problem here.
But Darksiders III, Furmark or Superposition, cannot (it results in GPU core clocks to fluctuate and produce very jaggy performance).

 

I tried superposition benchmark, and while the score seems ok more or less.
Here's the problem that occurs with fluctuating GPU frequencies (recorded the benchmark run with my phone) :


This exact GPU frequency fluctuation occurs in Darksiders III.
As I said, Star Trek Online and Luxmark seem fine as the GPU clocks on those 2 run consistently on 1077MhZ... but Superposition and Darksiders III are making GPU clocks jump all over the place.

 

Have GPU-Z and HWiNFO64 sensors open when running the problematic applications, and note what the GPU performance limit reason is when the clock speed dips.

 

It could be a possible VRM overheating issues... not quite sure...

 

Here's the HWinfo screen.
I took it after running the benchmark.
I'll have to hit the gym now and I'll try running another benchmark when I get back with GPU-Z active.
I'll make another video recording if necessary.

 

That doesn’t tell me anything, you would have to show the GPU limit reasons section during one of the performance dips.

 

How would I go about doing that?
Where is the GPU limit reasons section in HWinfo? I don't see it.


Also, the performance dip is constant throughout running the Superposition benchmark or Darksiders III... it doesn't occur at specific intervals... the gpu clocks are oscillating wildly every half a second ALL the time when the GPU is stressed by something like Superposition or Darksiders III.

Less demanding game such as Star Trek Online doesn't produce same gpu clock oscillation, neither does Luxmark.

 

Guess HWiNFO doesn’t have a performance limit reasons section for AMD GPUs. Maybe GPU-Z will show it.

Doesn’t Asus use that nasty thermal paste on the VRMs instead of thermal pads? It’s possible that the thermal paste has dried up, so the VRMs are overheating under strenuous GPU load.

 

they used thermal paste on everything. No thermal pads whatsoever.
It's my thought as well that this could be contributing to the issue... and if its not, its just a bad chip.

 

You mean these stats?



Granted I do have an nvidia GPU so.... *shrugs*

 

Btw... another user whose GL702ZC is of same config as mine has his RX 580 scored 40% higher in Superposition, with no oscillating clocks.
This is an issue confined to my GPU and/or cooling assembly.

 

sometimes game boosters can resolve bizzarre behaviours....try razer cortex or iobit turbo mode with advanced system care...could be a process interfering. if that doesnt work then its definitly temperatures

 

Run this program as admin and see if you can find older driver remnants that might be interfering with the device's performance.

 

I'll try, but I ran superposition with old drivers (I changed to the old ssd that came with the laptop) and the issue was the same. Those old drivers were installed by Asus repair centre.
Plus on my own ssd (before I switched it out) I used amd driver uninstall utility in safe mode to install newest drivers from the get go (doing so made me lose freesync because the laptop wants original drivers first. Then express install of newest on top of them to retain freesync function).

 

Have you tried Xtreme G modded AMD drivers?

 

No, besides, AMD's desktop drivers work just fine on my GL702ZC because it has desktop hardware to begin with.
So, why would I use Xtreme G modded AMD drivers?

Vengance is the user who ran Superoposition on his GL702ZC and he's getting 40% higher performance with no oscillating frequencies.
I very much doubt this is a software problem.

 

Discovered a problem.
GPU might be throttling and temperatures might not be reported correctly.
When I ran Superposition with fans set to Auto control, I got a score of 2500 (1080p and High details), whereas I should be getting close to 5000.
Temperatures under those circumstances reached 80 degrees Celsius, but the fans never ramped up from their 'low state'.

If I set fans to 80% manually, undervolt the GPU on the core to -93mV and set +50% power limit, I get 4862 score under same settings.
But, Meteeusz managed a score of 4966 in Superposition with his GL702C, and stock/default settings.

This raises the question on what the heck did Asus repair shop do to my unit to have messed up the cooling and GPU performance under high load so severely that I need to mod my GPU and set fans to manual so I'd get 'close' to what should be a 'stock score'?

Laptop is going back.
This time, I'm insisting on a full refund.
If the unit is incapable of producing stock GPU performance under auto-fan regulation (which should ramp up the fans to maximum under those circumstances, or close), then it has no business being out in the wild.

 

so it was temps...your sensors are shot.

edit if it makes you feel any better I had a 540 laptop and had to return it for similar reasons

 

Yes, but you're not dealing with stubborn idiots who keep insisting on repairing the unit as opposed to replacing it with a new one or giving you a refund.
I don't care Asus introduced a clause that says the unit needs to undergo 3 or 4 unsuccessful repairs before they offer a replacement or a refund (my guess is that Asus might have introduced this clause so they can keep 'repairing' the unit as long as they can and by the time it fails the 3rd time or 4th time, chances are the warranty would have expired anyway).
If that's the case, I'm not gonna fall for that. They have the consumer law to deal with now (which they are in clear violation of). I'm well within my rights to request a full refund given previous failures and the fact the laptop still isn't working properly.

 

A bit of update...
I created a formal letter based on Citizen Advice recommendation about consumer rights in the EU which is addressed to LaptopsDirect (trader, not the OEM) as they are the ones responsible (according to EU law) to deal with refunds and/or replacements (regardless of them trying to pass the replacement/refund to Asus - which won't give a relevant repeal number unless the unit underwent 3 or 4 unsuccessful repairs... and I think its absurd people should wait for so many RMA's before they can claim a replacement or a refund).
Considering that the trader has been passing me off to Asus, and as a result violated my consumer rights, I basically invoked the law on them, so they will have to deal with that now as they essentially sold me a defective unit that broke twice on its own, and still isn't working properly.

I asked for a full refund based off that (and yes, the unit is still under warranty, but that doesn't matter, as here in UK, the law is valid for consumer goods up to 5 years here in Scoltand - 6 years in England... I'm just 14 months into my warranty (out of 24 months).

You can check for details in this thread (and feel free to use the letter template if you encounter similar problems like I have) :
http://forum.notebookreview.com/threads/asus-rog-gl702zc-owners-lounge.809882/page-136#post-10838555

As I said, I'd love to open up the unit and try applying my own GC gelid Extreme therrmal paste to CPU and GPU and proper/high quality thermal pads on the VRM's and relevant chips to see if that solves the problem... but that would void the warranty (even though I should be technically allowed to do this as long as I am careful and don't break anything)... and seeing how this would have been a third RMA, and Asus Repair centre ignoring my requests to strain test the GPU/CPU to see how they and cooling assembly behave under those conditions, I figured enough is enough, and I've been more than patient.

I realise that in those 2 weeks, Asus repair centre might have dealt with the problem... but I don't trust this unit to come out of this whole debacle functional. And even if it does... how long until another potential breakdown occurs OUTSIDE the warranty period?
No, not having this hassle.

 

well done sir....I am hated by stores for returning defective laptops. and the truth is many of them are defective...i really do hope you are able to get a full refund via standing laws

 

I hope so too... after the debacle they put myself (and other people) through, I'm hoping I have enough on them to claim a full refund and then some if this ends up in small claims court.

Plus, the fact the retailers and OEM's are clearly ignoring EU regulations in EU and not informing consumers about their rights is obscene.
They should be sued for not mentioning anything about EU consumer rights laws alone (never-mind everything else) because it seems to me they are keeping people intentionally in the dark.

Also, I find it incredibly 'funny' how it takes seconds for a payment to clear for the company or government institutions, and if you (a consumer/regular citizen) need/request a refund, it takes 'up to a week' to get your money back.
From multi-million euro/£/$ companies/institutions?
I don't think so.

They get away with way too much.
So, let's see if consumer rights law can help me out.

 

What do you plan on purchasing if you get your full refund?

 

I would likely get an all amd laptop preferably (as I'd like to support AMD)… something like acer helios 500 if it's reliable.
If not, then I'd guess an equivalent intel/NV laptop (but prefer avoiding that if I can due to their bribing tactics - also. There's no guarantee thst an intel/nv replacement would be any more reliable)

Mind you, the concept behind all amd hardware is sound and efficient when done right, but Asus really dropped the ball on gl702zc in regards to cooling and possibly quality parts.

If no refund is possible, then I guess a functional replacement (that doesn't break down) will have to do (but if that happens, free extended warranty needs to be guaranteed).

I like the gl702zc when it works and the rx580 inside it is powerful enough... But these (Asus related) problems really ruined the experience for me.

 

Do you want a gaming laptop?
Check out Helios 500 forums before buying.

 

A productivity and occasional gaming laptop (though I suppose 6 cores running at higher frequency would do the trick too), plus I'm on sort of a budget... don't want to spend much more compared to what I did for GL702ZC (if I get a refund).
Acer Helios 500 as far as AMD hardware goes is only configured with 2700 and V56 (no rx 580 option)... and the thermals, noise, etc. look to be great (even under maximum load)... plus V56 and 2700 can both be undervolted to improve things further.

Price is also one of the reasons why I want all AMD hardware... it can usually be cheaper than Intel/NV option.

Not concerned too much with battery life since I don't plan on using the thing without a power outlet (a desktop isn't exactly a practical option in case of moving or going on a trip).