Radeon RX 5700 Series Thermal Issues Points To AMD and NOT AIB Partner Designs!

Nenu

[H]ardened
Joined
Apr 28, 2007
Messages
19,175
I have no wish to defend Asus but is there any evidence the cards that didnt fail used AMDs recommended pressure?
ie did other mfrs discover through their own testing that the pressure needs to be higher, and that is why they dont fail?
 

Derangel

Fully [H]
Joined
Jan 31, 2008
Messages
18,807
I have no wish to defend Asus but is there any evidence the cards that didnt fail used AMDs recommended pressure?
ie did other mfrs discover through their own testing that the pressure needs to be higher, and that is why they dont fail?
Even if AMD's recommendation was wrong (which is entirely possible) ASUS still should have properly tested the card before shipping it. Releasing a product with a problem that would have been easily caught with QA testing is still on ASUS' head.
 

Grimlaking

2[H]4U
Joined
May 9, 2006
Messages
3,191
Video cards are just tested if they power on. Aside from possible minor changes to the cold plate and placement of thermal pads, cooler and heatsink design is not changed for different video cards. The AIB partner would have no reason to not believe the specification guidelines provided by the chip manufacturer.

Shitting on ASUS for QA is an internet meme and considered acceptable. Putting just a little responsibility on AMD is taboo because they're considered the good guy underdog.
Let me ask you a question... and bare with me here... it might take some explaining.

For the 5700 chipset the cooler designed and tested by AMD had a specific setup/flow/handling of heat characteristics. If the default cooler setup works fine. How is AMD to know if a vendors custom cooler will work the same? Are they responsible for testing all of the possible custom cooler installtions to make sure that they work and define the spec for them on mounting and fan speeds and what not?

OR should the vendor when designing and implementing a CUSTOM cooling solution actually test it and not just say.. "40 lbs is 40 lbs."

When I put tires on my care that are outside of OEM spec, I don't go by the OEM tire pressure reccomendations any more. I go by the tires recommended spec. It's on me if they are under inflated or over inflated at the spec the manufacturer recommends. I have zero expectation that any tire that can fit my car was tested on my make and model of car, UNLESS it is the OEM spec for that car.

Sure in most cases it will be just fine. But the outside cases is where it counts.
 

Marees

Limp Gawd
Joined
Sep 28, 2018
Messages
344
I have no wish to defend Asus but is there any evidence the cards that didnt fail used AMDs recommended pressure?
ie did other mfrs discover through their own testing that the pressure needs to be higher, and that is why they dont fail?
I thought, Saphire works closely with AMD

They would have given feedback if something was not right

Or maybe it is standard practice for Saphire to go over & above AMD's recommendations
 
  • Like
Reactions: Auer
like this

Auer

[H]ard|Gawd
Joined
Nov 2, 2018
Messages
1,561
I thought, Saphire works closely with AMD

They would have given feedback if something was not right

Or maybe it is standard practice for Saphire to go over & above AMD's recommendations
Probably, and ASUS was pretty naive in this case. Sapphire knows AMD well.
 

travm

Limp Gawd
Joined
Feb 26, 2016
Messages
473
Mounting pressure eh. That's a complex subject. Did Asus actually measure the mounting pressure to verify they were meeting the recommended spec? Or just use borderline screws and give no shit?
The number of variables in screw clamp pressure are huge.
 

Auer

[H]ard|Gawd
Joined
Nov 2, 2018
Messages
1,561
Mounting pressure eh. That's a complex subject. Did Asus actually measure the mounting pressure to verify they were meeting the recommended spec? Or just use borderline screws and give no shit?
The number of variables in screw clamp pressure are huge.
Arent they spring loaded so as to not be able to over tighten and mounting force is determined by the spring strength?
 

Azrak

Gawd
Joined
Oct 4, 2015
Messages
906
Arent they spring loaded so as to not be able to over tighten and mounting force is determined by the spring strength?
Sure, but how many thermal pads are included under the mounting plate. How thick are they? Are they the same as the ones AMD used? Did AMD use any when speccing, or was the spec without any pads? Too many variables to know without being employed by AMD or one of the AIBs in question.
 

Nenu

[H]ardened
Joined
Apr 28, 2007
Messages
19,175
Even if AMD's recommendation was wrong (which is entirely possible) ASUS still should have properly tested the card before shipping it. Releasing a product with a problem that would have been easily caught with QA testing is still on ASUS' head.
All good and well but you didnt answer my question.
I am trying to determine whether AMDs recommendation has been used with success or not directly, there is no evidence of this yet despite the headline etc.
 

Derangel

Fully [H]
Joined
Jan 31, 2008
Messages
18,807
All good and well but you didnt answer my question.
I am trying to determine whether AMDs recommendation has been used with success or not directly, there is no evidence of this yet despite the headline etc.
It's a fairly irrelevant question, is it not? If AMD's recommendation was wrong then every single other AIB obviously realized this and made sure their cards had proper tension before release, except ASUS. AMD's recommendation was likely based on the reference model, which uses a thermal pad over the GPU die vs a thin layer of thermal paste that all AIBs use. If that's the case than the natural assumption would be that the AIBs would need to alter the tension as needed to ensure proper contact.
 

Nenu

[H]ardened
Joined
Apr 28, 2007
Messages
19,175
It's a fairly irrelevant question, is it not? If AMD's recommendation was wrong then every single other AIB obviously realized this and made sure their cards had proper tension before release, except ASUS. AMD's recommendation was likely based on the reference model, which uses a thermal pad over the GPU die vs a thin layer of thermal paste that all AIBs use. If that's the case than the natural assumption would be that the AIBs would need to alter the tension as needed to ensure proper contact.
Its directly relevant, was AMDs advice good with some AIBs or was it found to be defective by them all.
It prevents having to speculate as you just did.
 

Ebernanut

[H]ard|Gawd
Joined
Dec 15, 2010
Messages
1,146
Its directly relevant, was AMDs advice good with some AIBs or was it found to be defective by them all.
It prevents having to speculate as you just did.
I'm not sure how it's AMD's responsibility to make sure their recommended pressure works with custom cooling, especially since it doesn't sound like a requirement as much as a loose recommendation.

Also the fact that it wasn't an issue for everyone suggest that it was either an obvious issue to all the other AIBs or that it worked fine for them. I'm really not sure how they didn't notice unless they didn't even test their custom cooling on this card.
 

Lakados

[H]ard|Gawd
Joined
Feb 3, 2014
Messages
2,031
Who makes a waterblock for the Asus Strix 5700xt that's worth it?
EKWB makes a few that work with the Strix 5700xt, I have been happy with their products and support to date not 100% sure if it is worth it in the traditional sense but it if is worth it to you then by all means.
 

Nenu

[H]ardened
Joined
Apr 28, 2007
Messages
19,175
I'm not sure how it's AMD's responsibility to make sure their recommended pressure works with custom cooling, especially since it doesn't sound like a requirement as much as a loose recommendation.

Also the fact that it wasn't an issue for everyone suggest that it was either an obvious issue to all the other AIBs or that it worked fine for them. I'm really not sure how they didn't notice unless they didn't even test their custom cooling on this card.
I didnt suggest anything of the sort.

I question your second point because it is exactly what I have been asking.
Where did you get that as fact?
Was it really not an issue for some AIBs or did they find it an issue and correct it?
 

DogsofJune

2[H]4U
Joined
Nov 7, 2008
Messages
3,675
EKWB makes a few that work with the Strix 5700xt, I have been happy with their products and support to date not 100% sure if it is worth it in the traditional sense but it if is worth it to you then by all means.
I saw that one, hesitant to do another EK product after my issues in the past, but things change I suppose. The price though.... I already bought the "pricey" 5700XT

I see tidbit's about a bykski/Aliexpress waterblock but haven't discovered where it's sold from. Don't see it listed on their site.

Just trying to see my options with this piece of hardware. Don't really want to send it off, but I also don't want flaky performance.
 

Derangel

Fully [H]
Joined
Jan 31, 2008
Messages
18,807
I saw that one, hesitant to do another EK product after my issues in the past, but things change I suppose. The price though.... I already bought the "pricey" 5700XT

I see tidbit's about a bykski/Aliexpress waterblock but haven't discovered where it's sold from. Don't see it listed on their site.

Just trying to see my options with this piece of hardware. Don't really want to send it off, but I also don't want flaky performance.
Bitspower and Alphacool both have blocks as well.
 

Nenu

[H]ardened
Joined
Apr 28, 2007
Messages
19,175
I thought, Saphire works closely with AMD

They would have given feedback if something was not right

Or maybe it is standard practice for Saphire to go over & above AMD's recommendations
The point is to remove the guesswork otherwise the arguments have little foundation.
 

Ebernanut

[H]ard|Gawd
Joined
Dec 15, 2010
Messages
1,146
I didnt suggest anything of the sort.

I question your second point because it is exactly what I have been asking.
Where did you get that as fact?
Was it really not an issue for some AIBs or did they find it an issue and correct it?
Clearly not everyone is having issues with their custom cooling which is all I stated as fact. It does sound like the other cards with cooling issues were likely related to something else though which would make Asus the sole one with this particular issue.

I don't really think it matters whether it wasn't an issue for others or if they fixed it because custom cooling should be properly tested on every card. My main point was that it's up to the AIBs to make sure their cooling solution works and following some guideline doesn't absolve them of that.
 

Jim Kim

2[H]4U
Joined
May 24, 2012
Messages
3,700
Clearly not everyone is having issues with their custom cooling which is all I stated as fact. It does sound like the other cards with cooling issues were likely related to something else though which would make Asus the sole one with this particular issue.

I don't really think it matters whether it wasn't an issue for others or if they fixed it because custom cooling should be properly tested on every card. My main point was that it's up to the AIBs to make sure their cooling solution works and following some guideline doesn't absolve them of that.
This isn’t rocket science, well maybe for ASUS it is.;)
Whodathunkit a kilograms worth of copper, aluminum and 3 fans would require a different mounting solution than an oem blower unit.
 

Imhotep

Gawd
Joined
Feb 12, 2014
Messages
815
My Strix 580's all have a 3/16 of thermal padding on their VRMs. What is the point of the 3 fan big ass heatsink if you cover the hot parts with such a thick thermal pad.
Nice try ASUS
 

Nenu

[H]ardened
Joined
Apr 28, 2007
Messages
19,175
Clearly not everyone is having issues with their custom cooling which is all I stated as fact. It does sound like the other cards with cooling issues were likely related to something else though which would make Asus the sole one with this particular issue.

I don't really think it matters whether it wasn't an issue for others or if they fixed it because custom cooling should be properly tested on every card. My main point was that it's up to the AIBs to make sure their cooling solution works and following some guideline doesn't absolve them of that.
Yes but that is not what I was talking about.
I wonder if any AIBs who followed AMDs advice created a working solution without flaw. This appears to be unknown despite the thread title.
 

Pantalaimon

Limp Gawd
Joined
Aug 17, 2006
Messages
191
Yes but that is not what I was talking about.
I wonder if any AIBs who followed AMDs advice created a working solution without flaw. This appears to be unknown despite the thread title.
Are you referring to working custom cooling solution, or reference? If it's about the working custom cooling solution, then it seems Powercolor, or Sapphire custom cards do not have issues. We don't know if they followed AMD's guidelines, or went above and beyond those guidelines. So, there two scenarios:

1. If they did go above and beyond, then it seems they were more strict in their design and QA than Asus, since their cards don't have the same issue as Asus' cards
2. If they only followed AMD's guidelines for their custom cards, but still don't have the same issue as Asus does, then it seems Asus has a manufacturing and QA issue.

Either way it doesn't paint Asus in a good picture, although you could also fault AMD in the first scenario for not checking up on their partners' designs and alerting them that maybe the guidelines are not enough, when it's a custom design.

- Vu
 

M76

[H]F Junkie
Joined
Jun 12, 2012
Messages
10,480
Just because there is a recommended mounting pressure doesn't mean it will work for all coolers and designs exactly the same.
Car manufacturers have recommended tyre pressures too, but you change them when you fit different size wheels.
This is exactly the same situation, you fit different coolers with different characteristics on a custom PCB, and don't make sure the cooling is actually adequate is a huge oversight. Or perhaps calculated risk to save on QA.
 

Snowdog

[H]F Junkie
Joined
Apr 22, 2006
Messages
10,232
Just because there is a recommended mounting pressure doesn't mean it will work for all coolers and designs exactly the same.
Car manufacturers have recommended tyre pressures too, but you change them when you fit different size wheels.
This is exactly the same situation, you fit different coolers with different characteristics on a custom PCB, and don't make sure the cooling is actually adequate is a huge oversight. Or perhaps calculated risk to save on QA.
HWUB rips Asus a new one:
 
Joined
Apr 1, 2008
Messages
2,604
Incredibly disappointed with ASUS, had a 5700 XT from ASUS, the TUF version to be precise and it turned out to be a massive disappointment. Card ran too hot at load (85C) and one time I remember it overheated and the entire system shut down because of it.

ASUS as of late is a huge disappointment. I've had many problems with their products I've purchased the last few years and I will be avoiding them from now on.
 
Joined
Apr 11, 2017
Messages
540
It almost sounds like ASUS hired a person with a lot of responibility that doesn't know what they are doing and trying to cut corners.
 

sabrewolf732

Supreme [H]ardness
Joined
Dec 6, 2004
Messages
4,193
So Steve from hwub/techspot put up a video on this
Seems like it's really an Asus problem. The heatsink is far too large/heavy for the number of screws they utilize, the heatsink doesn't even make contact with the vrms.

Sorry Asus.

Thread title is incorrect imo.
 

Derangel

Fully [H]
Joined
Jan 31, 2008
Messages
18,807
So, after their statement was torn to shreds across the internet ASUS decided to change the wording of their post.

Former statement:

Initial batches of ROG Strix 5700-series graphics cards were built following AMD’s guidelines.
New statement:

Initial batches of ROG Strix RX 5700-series graphics cards were torqued to 30-40 PSI based on AMD’s baseline recommendations. While those guidelines provided leeway to apply more torque, we took a cautious approach because were dealing with a new 7nm GPU and didn’t want to risk damage to the die.
They have also completely removed the graph.

So, instead of properly testing their cards they went by the baseline, even though AMD said they could apply more pressure, and shoved it out the door without ensuring it would work properly.
 

Snowdog

[H]F Junkie
Joined
Apr 22, 2006
Messages
10,232
I am not cutting Asus any slack, but this again reminds me how the old ATX form factor is creaking under the load of modern high power components.

2-3lbs coolers hanging off of a card in a slot isn't a great idea, even when nothing goes wrong.
 

Derangel

Fully [H]
Joined
Jan 31, 2008
Messages
18,807
I am not cutting Asus any slack, but this again reminds me how the old ATX form factor is creaking under the load of modern high power components.

2-3lbs coolers hanging off of a card in a slot isn't a great idea, even when nothing goes wrong.
Outside of every case coming with support brackets I'm not sure if there's a good solution. The from factor is so entrenched now I don't know if there really could be a way to get everyone on board with changing it. Vertical mounts could become more standard, but they have their own host of problems (especially as most riser cables are garbage).
 
  • Like
Reactions: Auer
like this

Auer

[H]ard|Gawd
Joined
Nov 2, 2018
Messages
1,561
Outside of every case coming with support brackets I'm not sure if there's a good solution. The from factor is so entrenched now I don't know if there really could be a way to get everyone on board with changing it. Vertical mounts could become more standard, but they have their own host of problems (especially as most riser cables are garbage).
I'm all for vertical mounting, and it would seem that making a quality riser cable should not be all that difficult to achieve and could be included with at least mid/high end GPU's.
 

Meeho

Supreme [H]ardness
Joined
Aug 16, 2010
Messages
4,835
If you actually read what ASUS says AMD provided guidelines for mounting pressure in the range of 30-40 PSI, which is what early ROG batches followed. Let's not automatically assume the White Knight position for AMD whenever an issue for them pops up.
I have no wish to defend Asus but is there any evidence the cards that didnt fail used AMDs recommended pressure?
ie did other mfrs discover through their own testing that the pressure needs to be higher, and that is why they dont fail?
Increased pressure is a bandaid fix for Asus' bad cooler design. It's not AMD's pressure spec that is the issue, but Asus' design and mounting points.
 
Top