Latest C7H BIOS

[EZ_PC_TECH]

1. Version 2901 have some real issue with overheating,
CPU PWM & Q-FAN PWM (CPU & PCH) are ignoring critical temps over 75c, system is going to shut down by its own reaching 95c.

2. there is as well some strange read-out trouble with HWiNFO, CPU Socket Temp is jumping between 127c and 255c for some seconds once an hour, fans are going crazy.

My system RM650, 3900x, 2070s, CH VII

You have an issue with our cooling solution, not with a BIOS. I'm on 2901 and my temps not exceeding 64C under the AVX stress test

[RNG_AGESA]

You have an issue with our cooling solution, not with a BIOS. I'm on 2901 and my temps not exceeding 64C under the AVX stress test

its not the question of cooling solution (Noctua nh-d15), which is enough for undervolted 3900x.
5 case fans + 2 CPU fans are ignoring critcal temp (on C7H everything above 75c), not spinning with 100% as they should.
air control is broke i assume. it was same on 2606, 2703, 2901 - exception 2801.

[EZ_PC_TECH]

its not the question of cooling solution (Noctua nh-d15), which is enough for undervolted 3900x.
5 case fans + 2 CPU fans are ignoring critcal temp (on C7H everything above 75c), not spinning with 100% as they should.
air control is broke i assume. it was same on 2606, 2703, 2901 - exception 2801.

What about the thermal compound application? The old-school "blop in the middle" doesn't work for ryzens (it does overall, but temps are significant higher).

[RedSector73]

What about the thermal compound application? The old-school "blop in the middle" doesn't work for ryzens (it does overall, but temps are significant higher).

ROTFLMAO any method is fine. I prefer a layer spread over the chip but apart from not using TIM this is total crap. Method does not matter that much.

[EZ_PC_TECH]

ROTFLMAO any method is fine. I prefer a layer spread over the chip but apart from not using TIM this is total crap. Method does not matter that much.

You are referring to outdated video from 2016, in which was tested a CPU with a single chip which is not the case with Zen2.
Application DOES matter now, and I know it very well, as well as the amount of paste to 1/10 mL, exact pressure, and method that yields the best results. That's why I have 64-65C under any torture test, while you are overheating on undervolted CPU and exact the same BIOS.

[Lothos]

You are referring to outdated video from 2016, in which was tested a CPU with a single chip which is not the case with Zen2.
Application DOES matter now, and I know it very well, as well as the amount of paste to 1/10 mL, exact pressure, and method that yields the best results. That's why I have 64-65C under any torture test, while you are overheating on undervolted CPU and exact the same BIOS.

Wouldn't it be easier to just use this instead of thermal paste from now on?

https://www.amazon.com/gp/product/B0...SO4PYH3S&psc=1

Been researching it and seems people have been reporting better results than paste and it is reusable.

[EZ_PC_TECH]

Wouldn't it be easier to just use this instead of thermal paste from now on?

https://www.amazon.com/gp/product/B0...SO4PYH3S&psc=1

Been researching it and seems people have been reporting better results than paste and it is reusable.

Not really. I suspect that reviews that you see are paid. We have an overclocking team here and tested pretty much any thermal paste/pad available on the market and this pad as well. This pad isn't terrible, but good thermal paste will give you better results since it can get in every microcavity of the contact surfaces. My favorite paste is Thermalright TF8, it hands down better than all Noctua, Cryorig, MX, etc pastes I tried. Just get it (only $10) and apply it this way: https://imgur.com/a/l36QZES
One regular blop in the middle and 4 smaller in the corners. This way you'll cover IOD and CCDs entirely on any Ryzen. You'll need 0.1 - 0.11mL of thermal paste to cover the entire IHS.

[RedSector73]

Not really. I suspect that reviews that you see are paid. We have an overclocking team here and tested pretty much any thermal paste/pad available on the market and this pad as well. This pad isn't terrible, but good thermal paste will give you better results since it can get in every microcavity of the contact surfaces. My favorite paste is Thermalright TF8, it hands down better than all Noctua, Cryorig, MX, etc pastes I tried. Just get it (only $10) and apply it this way: https://imgur.com/a/l36QZES
One regular blop in the middle and 4 smaller in the corners. This way you'll cover IOD and CCDs entirely on any Ryzen. You'll need 0.1 - 0.11mL of thermal paste to cover the entire IHS.

Can you stop with your unproven method of ditching everything based on your own limited exposure to Ryzen. How many of these systems do you own or have you built ?

You seem quite young or naive or both. Explain this then.

[RedSector73]

And this


I have better temps than you must be better at TIM than you - or could it be something else cools my cpu and it just does heat transfer between two interfacing surfaces ?

[Frostie36]

I must admit seeing the post earlier on here about a pea size in the middle not working for ryzen is rubbish imo as that's the application method I used & my temps are not outrageous on my 3900x using a noctua nh-u12s cooler, I also used the noctua tim that came with the cooler, not to mention there's barely any difference in temps between all the high end thermal interface products out there, 1-2c at most as demonstrated by various videos I've seen on tech channels on youtube.

The chiplets aren't in the middle like a monolithic die but they're nowhere the edges either so the usual application method of applying tim is fine on ryzen imo & going by my own experience as the pressure of the heatsink will spread the paste out to cover i/o die & chiplets.