Clevo P775 power limit?

[tway]

I don't know how old these are but when I take my heatsink off my thermal pads have noticeable physical indents from the components they're contacting (similar to yours in the bottom left). Maybe you have the same issue as DoenerBoy where yours have minimal/no proper contact due to not being thick enough. I could be wrong or not seeing properly from the pics but either way replacing them will be much cheaper than any actual hardware so may as well try imo. Also is that liquid metal on your die?

[Kniben]

@tway First of all, congratulations on fixing your GPU! As for my heatpads, it's probably not a bad idea to replace them. Also I need to check those screenshots again closely, if all relevant components are covered. As for the indentations, sure, flipping the pads should solve that. It might be liquid metal, has a golden colour. Don't worry, I used a modest amount and there is plenty of insulation around the die, so that won't cause a short circuit.

 

Edit: Did some research on liquid metal and what they mean is Gallium, that has a silvery colour, so no, it's not liquid metal 🙂

 

Now, regarding temps, if Furmark doesn't crash, I left it for about half an hour (-450 core increment):
 

GPU: 80°C

GPU hotspot: 93°C

VRAM temp: 90°C

 

How are your temperatures with the new heatpads?

[1610ftw]

2 hours ago, tway said:

Yeah it's unfortunate the P775 has these design flaws, but also I find your experience with Eurocom crazy... 

 

I personally never bought from them since they looked like scammers to me, but for the prices they're charging I expect them to fly out a representative to your doorstep as customer support 😂 who do they think they are charging this much? Apple? They are very clearly trying to exploit us, I have talked to so many private sellers with overpriced Clevo parts who point to the Eurocom ebay shop as a justification. And this is their goal all along, they know exactly what they're doing. Best we can do is not buy from them, I don't know how they aren't out of business or anything by now, some of their items have been sitting for over 2-3 years with no sale. 

 

Anyway this is off topic now but hopefully this dissuades people from giving those scalpers any money. It wasn't always like this but as of present that company is doing much more damage to the community than adding anything of value. @Kniben I hope it's just your thermal pad placement that's bad so you don't have to buy a new motherboard, especially from them. I've attached images I found in some threads that I used as references to check thermal pad placement, hopefully these help.

 

 

 

 

 

Last thing about Eurocom from me: In the end it was almost comical to see how they weasled themselves out of doing the right thing. It was not about a big amount of money and I shudder to think what would have happened if I had bought something really expensive from them.

 

Looking at your GPU it is clear that paying attention to the pad thickness and choosing the right pads seems to be quite involved and makes me consider to finally try my thermal putty that I have sitting arround for quite some time now - if only it wasn't so messy.

 

 

 

 

[Kniben]

@1610ftw You asked if somebody could use an E-meter to measure actual power draw from the AC side. Here are the results, mind you (-450 core increment)

Furmark 3840 x 2160: 198 W

prime95 Small FFTs, all 8 cores: 194 W

Both of the above: 322 W

 

The thermal pads I used were stock and I used them since I bought the thing in 2019. I doubt the technicians at XMG did a poor job during assembly, but I noticed one of the thermal pads not covering the inductor at the bottom and that tiny square chip. So perhaps the coil just died. I also don't know if the inductors are staged according to core frequency, meaning not all of them are used, if the core frequency is set to a lower limit. I can achieve stability by reducing the core increment by -450, while tway needed -700. Heck, I don't know how the GPU regulates its power, all I know is, when things get hot, they break or at least degrade over time and I think that has happened with the coil I mentioned. Perhaps the GPU is unsalvageable and I should buy a replacement.

[1610ftw]

51 minutes ago, Kniben said:

@1610ftw You asked if somebody could use an E-meter to measure actual power draw from the AC side. Here are the results, mind you (-450 core increment)

Furmark 3840 x 2160: 198 W

prime95 Small FFTs, all 8 cores: 194 W

Both of the above: 322 W

 

The thermal pads I used were stock and I used them since I bought the thing in 2019. I doubt the technicians at XMG did a poor job during assembly, but I noticed one of the thermal pads not covering the inductor at the bottom and that tiny square chip. So perhaps the coil just died. I also don't know if the inductors are staged according to core frequency, meaning not all of them are used, if the core frequency is set to a lower limit. I can achieve stability by reducing the core increment by -450, while tway needed -700. Heck, I don't know how the GPU regulates its power, all I know is, when things get hot, they break or at least degrade over time and I think that has happened with the coil I mentioned. Perhaps the GPU is unsalvageable and I should buy a replacement.

 

Thanks for taking those measurements - at least 322W is no problem for the power supply provided that it is not defective..

 

Your temps seem quite hot and I suggest you get those temps sorted out first, as with a total power uptake of less than 200W your GPU probably only consumes 150W and it should be a lot cooler at 150W.

 

With a card that old I would shoot for 70C max for the GPU and 85C for the VRAM when gaming and then take it from there and replace only when these conditions have been met and it is still acting up.

 

If you are not sure what to do you could either improve the existing pads and add some PTM7950 or you could get some thermal putty which is a bit more forgiving but also more of a mess:

 

 

 

 

 

[tway]

@Kniben my GPU still works great under full load even after extensive testing, still no power cuts so far. Have you tried replacing the thermal pads in accordance to the diagrams I sent and checking that they're making good contact with the heatsink? Maybe you're making the same mistake I made, it is quite hard to tell without any references. The fact you only need to underclock to -450 to be stable whereas I needed -700 (due to a missing thermal pad altogether) tells me perhaps you have them all in correct positions but maybe they lack proper contact with the heatsink especially since you mentioned they're incorrect thickness, hence poorly dissipating heat

[Kniben]

@1610ftw You are right, I should try to reduce the temperatures. I have run another combined CPU and GPU benchmark, so I'm pushing the cooling system to its maximum: GPU core: 83°C, GPU hotspot: 95°C and VRAM: 93°. GPU power draw is at 140 W, due to the reduced clock.

Also what I did was play around with the core increment again, and it still remains stable at -450 (I can start and stop Furmark indefinitely) and as soon as I set it to -445, I can start and stop the benchmark perhaps three times before it shuts off again. So now we have a definitive reproducible scenario. I still think that my issue arises from the EC deciding to activate an additional MOSFET at -445, which is probably not performing well and causes the system to shut off. But without an expert here, I am unable to prove this. 

 

I have decided I will order some thermal heat pads and replace them. It's a bit unclear which thickness I should use for the different coil arrays. @tway Perhaps you can clarify? Given, that we use the exact same heatsink assembly: 2.0 mm for the VRAMs, 1.0 mm for the 4 coils, 3.0 mm for the FETs directly adjacent to them. 1.5mm for the 7 coils at the bottom, 1.0 mm for the FETs underneath, since the pads are resting on a "balcony" on the heatsink. What did you use for that small square chip next to the 4 VRAMs? 1.5 mm or 1.0 mm? I'd say go for 2.0 as that would be even with the VRAMs.

 

Sorry for me being so thorough 🙂 

 

[1610ftw]

Hope you get to the bottom of this as your GPU temps at only 140W do indeed seem excessive.

 

Can you check which temperatures you end up with in Time Spy?

In my experience its max temps are a better indicator for temps you may find in gaming and I am curious how much lower they would be than your Furmark temperatures.

 

Just checked a run I did with an XMG P775 with the 9900K and it went up to 62C tops with an almost 11K score but if I had to guess it is possible that I kept the bottom cover off and I always prop up the back by about 5cm with a folding stand.

 

 

[tway]

@Kniben Hey sorry, which chip do you mean exactly? There's a few in that location, I believe some don't have any pads covering them on mine

[Kniben]

9 hours ago, tway said:

@Kniben Hey sorry, which chip do you mean exactly? There's a few in that location, I believe some don't have any pads covering them on mine

This one here

 

What about the 1RO inductor and chip on the other side? Wouldn't it make sense to cool them as well? They aren't marked at all in the earlier images.

[tway]

I believe they aren't marked because cooling shouldn't be necessary for them, not sure but I don't have thermal pads on mine either and it's fine