• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

The Radeon RX 5700 XT Owners Thread.

Yeah, the reason for the slightly over spec 850w PSU was so I could connect the 2 separate PCIe cables as well as the 2 separate 8 pin and 4 pin EPS cables to the board, ruling out any potential issues due to splitting or daisy chaining. It also means I’ve got enough juice if I throw in the towel and chuck a 2080 or similar in there in frustration.
 
Yeah, the reason for the slightly over spec 850w PSU was so I could connect the 2 separate PCIe cables as well as the 2 separate 8 pin and 4 pin EPS cables to the board, ruling out any potential issues due to splitting or daisy chaining. It also means I’ve got enough juice if I throw in the towel and chuck a 2080 or similar in there in frustration.
Right now it does seem that 19.10.1 has improve on stabilizing gpu frequencies in the boost ranges.
I can only hope that with 19.10.2 it's further refined.
 
On idle the fans do not spin. And if on idle the card hits 60C means you do not have good positive airflow in your system. Any card with zero-fan speed on idle would experience the same issue here.
You can go then to fan setting in wattman and set the minimum speed to 10%. The card will continuously run the fans.

And add some extra fans to improve the positive airflow in the case (positive airflow means incoming fresh air)
The problem was that under 19.9.3 the temps were fine at idle, about 44 degrees. Only when running under 19.10.1 did the idle temp rise to 60. If this problem didn’t exist then I’d be using 19.10.1 happily as it fixed the boost issue.
 
Here here!
Although some believe in CPU bottlenecks, and they can happen, it's really the drivers which tell the cpu what to do in 3d gaming, full stop! If your CPU utilization isn't reading 95% and higher during gaming you are not cpu restrained.

That in my opinion is just not correct. There are plenty of scenarios where the cpu won't even reach 60% utilization(quads with ht/smt and up) and still be the main bottleneck. Applications/games that focuses heavily on few threads is a excellent example of this. That has nothing to do with the drivers and the entire statement is flawed imho.
 
The problem was that under 19.9.3 the temps were fine at idle, about 44 degrees. Only when running under 19.10.1 did the idle temp rise to 60. If this problem didn’t exist then I’d be using 19.10.1 happily as it fixed the boost issue.

If you are using AIB card, most have zero fan speed at temps bellow 60C. If you are so concerned raise the minimum default fan curve point to 10%
 
Unfortunately the safe mode DDU and ccleaner runs haven’t improved things. With my 2 NF-A14 front intake fans at the previous 32% which kept my 19.9.3 idle at 44 degrees the temp rises to 60 at which point the card fans kick in, reducing the temperature back under the 0% fan threshold before switching off. If I run the front intakes at 100% I can keep the card at 54 degrees but that’s not a solution.

Nothing other than the offline installer for 19.10.1 was run, and I’ve not installed Relive and made sure Enhanced Sync is permanently off. TRIXX 7.0 is installed and all I’ve done with that is switch BIOS modes.
 
That in my opinion is just not correct. There are plenty of scenarios where the cpu won't even reach 60% utilization(quads with ht/smt and up) and still be the main bottleneck. Applications/games that focuses heavily on few threads is a excellent example of this. That has nothing to do with the drivers and the entire statement is flawed imho.
You're assumption is flawed and a tired one to say the least. You base it under the assumption that CPU usage is a general average of cores based on game(s) that only use 1-2 cores (for example). And strawman that to ignore the implication of each individual core in a cpu (which I was referring to). And you are absolutely wrong to advise anyone that they cannot see a high cpu utilization as a means to determine if a gpu is being restrained by the cpu. To even imply it is asinine.
 
Last edited:
If you are using AIB card, most have zero fan speed at temps bellow 60C. If you are so concerned raise the minimum default fan curve point to 10%
That seems to be the only option at the moment, but then the in game fan rpm seems to be excessively high compared to previously. I’ll keep on testing out the various methods people have kindly suggested.
 
I might get the SoTTR demo and see if that’s better with it being a native DX12 title, but the inconsistencies are the things that really frustrate me. I’d sort of accepted that I may have to deal with the higher idle temps if it meant I was getting the performance out of my card, but if that fails as well...
 
You're assumption is flawed and a tired one to say the least. You base it under the assumption that CPU usage is a general average of cores based on game(s) that only use 1-2 cores (for example). And strawman that to ignore the implication of each individual core in a cpu (which I was referring to). And you are absolutely wrong to advise anyone that they cannot see a high cpu utilization as a means to determine if a cpu is being restrained. To even imply it is asinine.

Wauw, you took that very poorly. The fact that you make such a simpleminded statement and claim that a cpu has to be running at 95%ish before it can be considered a bottleneck is the only poor thing to write. I'm not saying that drivers cant be at fault for poor utilization of a cpu, of course they can, but it isn't that simple all the time. It's never just this one thing(drivers being bad or poorly done) that is the entire and only answer. Furthermore I just gave an example of one potential issue with your statement and I find it astounding that you from that single example can conclude what you did. If anything it just shows that you aren't open for criticism and clearly didn't understand what I wrote.

I stand by what I wrote earlier, you are wrong to claim that a CPU is never a bottleneck or part of a bottleneck in any or all situation just because CPU usage isn't showing 95+%. It is simply much more nuanced than that.
 
Wauw, you took that very poorly. The fact that you make such a simpleminded statement and claim that a cpu has to be running at 95%ish before it can be considered a bottleneck is the only poor thing to write. I'm not saying that drivers cant be at fault for poor utilization of a cpu, of course they can, but it isn't that simple all the time. It's never just this one thing(drivers being bad or poorly done) that is the entire and only answer. Furthermore I just gave an example of one potential issue with your statement and I find it astounding that you from that single example can conclude what you did. If anything it just shows that you aren't open for criticism and clearly didn't understand what I wrote.

I stand by what I wrote earlier, you are wrong to claim that a CPU is never a bottleneck or part of a bottleneck in any or all situation just because CPU usage isn't showing 95+%. It is simply much more nuanced than that.
You are rambling. And your 95% strawman want's it's 5% back.:D

I didn't discuss with you about drivers, per say, but to correct your assertion that one cannot use MSI AfterBurner, Hardware Info and other utility apps to provide insight as to a cpu issue in games. Which is absolutely incorrect.

To further add, one can research and find out if a game uses more then 2 cores (like BFV) vs a game that doesn't and know if it's even necessary to see individual core usage.

But yeah this whole thing really looks to me to be an issue with how the drivers keep the gpu feed. The low gpu utilization maybe an indication of that even though the clocks lower as well.
Although some believe in CPU bottlenecks, and they can happen, it's really the drivers which tell the cpu what to do in 3d gaming, full stop! If your CPU utilization isn't reading 95% and higher during gaming you are not cpu restrained.
This statement is your trigger and it has nothing to do with your 95% claim. It has everything to do with you not understanding that for the 5700xt low gpu utilization/low boost bug doesn't relate to cpu bottleneck.

Let be clear here. If there was a recommended CPU for the 5700xt AMD would have provided it in the RX 5700XT spec. portion of the page.
https://www.amd.com/en/products/graphics/amd-radeon-rx-5700-xt
There requirements involve the PSU. Take note of that. If there is a limit to what cpu a person can use in order to "fully utilize Navi" that information must come from the source, AMD, not you.

If you actually took the time to read some of those links I provided you will know that owners of the 5700xt have been doing cart wheels trying to get their card to work correctly.
-reinstall OS to win10 1903
-reinstall games
-upgrading power supplies to ultra/platinum, gold certs.
-uninstall/reinstall drivers
-go on forums asking for help
-tweaking wattman
-etc

When a person exhaust all avenues trying to fix something that wasn't broke with their old video card they tend to think there is issue with the hardware or driver for the hardware. Don't take it personal. Just put on one of the shoes that most of these frustrated owners are having and walk that mile with them. Show or find out how to empathize with those who are having issues.
 
Last edited:
Last one from me tonight to update that I’ve just rebooted using only my old monitor and the card temps are sitting at 44.
The latest drivers may not like my (admittedly) janky setup of using my old iMac in target display mode whilst I wade through the quagmire of getting a decent 27” screen. I’ll do some more tests tomorrow to see if I can confirm this but it’s reassuring that it may not be a hardware issue with the card.
 
Unfortunately the safe mode DDU and ccleaner runs haven’t improved things. With my 2 NF-A14 front intake fans at the previous 32% which kept my 19.9.3 idle at 44 degrees the temp rises to 60 at which point the card fans kick in, reducing the temperature back under the 0% fan threshold before switching off. If I run the front intakes at 100% I can keep the card at 54 degrees but that’s not a solution.

Nothing other than the offline installer for 19.10.1 was run, and I’ve not installed Relive and made sure Enhanced Sync is permanently off. TRIXX 7.0 is installed and all I’ve done with that is switch BIOS modes.

Could it be TRIXX is doing the same issue as Afterbuner and conflicting with Wattman?

NVM, read your last post about the monitor. But, i still suggest not using TRIXX or Afterburner.
 
One dead red devil card here, I'll confirm today, comp restarted late last night and no video output, lights work, fans don't spin - not sure they ever did during startup but..... Tried card in other comp same thing, tried other card in this comp and it all works.

I went amd cos I didn't want an rtx death :s lol
 
I’m not going to bother with Afterburner but TRIXX has the BIOS changing facility that’s very useful. It also apparently doesn’t do anything else in this new version with the 5700 XT as there are none of the functions in the older versions, apart from the RGB customisation.
 
Back
Top Bottom