• 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.

Dam Thought I had a 5700 xt for £370 from a non scalping retailer to tide me over until the 6800 settles in price but just beaten to it. If you're not fast you're last.
 
Have a 5700XT Nitro +

Have a really weird issue. Basically on the early drivers (before the Adrenaline ones came out iirc) no issues.

However since the Adrenaline drivers came out (the redesigned software iirc) whenever I use the PC from cold boot the GPU doesn't make a peep in terms of coil whine, nothing unless I'm pushing stupid high FPS in menus etc.

But when I wake from sleep overnight or after a good while in sleep the card makes a racket in terms of coil whine, just moving the cursor round the screen or scrolling in a browser, it's louder than even when I'm getting coil whine from high FPS.

It's taken me ages to narrow it down, it is definitely the GPU, it never did it when I first got it, it seems to occur after the updated Drivers last year and it only happens when waking from sleep. Have tried turning everything on/off in terms of power saving in the BIOS and windows but it always coil whines really bad only after waking from sleep.

Anyone else get really bad coil whine from the 5700XT after waking from sleep? I believe it's S3 sleep, only sleep mode enabled on my Windows install.
 
ShiWarrior said:
mines at 1742 for some reason
Click to expand...
Been irking me too. I assumed one chip was running hot so tore it down, liquid metal'd the gpu and applied some more pads where needed on the memory chips (MSI 5700 XT Gaming X).

Have to assume guy i bought it from dropped it or something as the upside is a 35 degree drop in Hotspot and now it holds boost rock steady in benchmarks and stresstests.

Never did find any reason for the drop in mhz, can't remember the split exactly but there could be a higher memory strap at 1750 so running just below runs faster? I know on my card (hynix chips) there is definitely a new timing strap at 1875 that meant it was slightly better running it about 1872.

EDIT: Keep seeing this in everyone's submissions for various 5700 xt's

GPU VRAM: 8176 MB @ 1742 MHz

Had a look on T.P.U. bios database, here are the timings:

Memory Timings (Samsung) tCL
125 MHz: 5
250 MHz: 5
500 MHz: 8
750 MHz: 12
875 MHz: 14
1000 MHz: 16
1250 MHz: 20
1550 MHz: 24
1800 MHz: 28
2000 MHz: 32
2250 MHz: 35

Memory Timings (Micron) tCL
125 MHz: 4
200 MHz: 4
500 MHz: 8
750 MHz: 12
875 MHz: 14
1000 MHz: 16
1250 MHz: 20
1550 MHz: 24
1750 MHz: 28
1800 MHz: 28
1875 MHz: 30
2000 MHz: 32

So possibly they are running Micron below 1750 for a better tCL. Interesting.
 
Last edited:
Tired9 said:
I sold my 5700 XT for more than I paid for it thanks to the crazy prices and sealed a deal on a GTX 970 for very cheap, overclocked the GTX 970 to 1550mhz and run games with absolutely zero problems, so blissful.

The Gigabyte Aorus 5700 XT was pure dog poop and 100% I am not going back to AMD ever again, it's not the case that yes the chances are they will have another lineup without these problems (RX 480 and 570 I have had before were excellent) but because they fluctuate so hard on drivers every lineup that is released, It's not worth my time or sanity.

I owned the card for over 9 months and finally gave up.

People will point at the person having the issues rather than accepting that simply, it's all **** and should not be happening.

For now I am happy with my 970, I am aiming to pick up a 3060Ti in time.
Click to expand...

I havent had a single problem with mine, upgraded from the R9 290(the same speed as a 970) a year ago and it just works;)
 
I can introduce loads of problems but at stock its been fine if I stayed on recommended drivers ;)

With my MSI RX 5700 XT Gaming X, messing about some more with memory speed. Used RBE and MPT to set max VRAM speed to 2100 on the slider instead of the 1900 it can't consistently do on its Micron chips anyway. Also set 50% power and max core to 2175 just in case more power meant more core headroom (It didn't)
Mainly I'd read somewhere VRAM can get stable again above the 2000 mhz mark with the looser timing strap.

After the bios flash asked Radeon what it would suggest for an AUTO OC on the VRAM now it has some (admittedly unexpected in its algo's) headroom. 2060mhz it suggested which seems a bit much lol. Something to aim for later!

Currently benching and stressing at 2020 mhz set and getting solid 2012mhz. Will have to run it through a few games now, checking how it affects maximum core.
 
Advice please, bought a 2nd user reference edition rx 5700 xt, the original owner said they repasted the card, but based on watching the temps on the core hit 90 degrees as opposed to watching and reading reviews for the reference edition the core should be way closer to 77 to 80 degrees, clearly they have done a bad job of repasting the gpu, I do know the ref edition uses Hitachi thermal pad, so if repasting with thermal paste , then four nylon washers are recommended to add to the mounting pressure.

I am going to use Noctua H2 paste.

I was just wondering if anyone else in the thread has repasted their reference edition rx 5700xt?
 
justwondering said:
Advice please, bought a 2nd user reference edition rx 5700 xt, the original owner said they repasted the card, but based on watching the temps on the core hit 90 degrees as opposed to watching and reading reviews for the reference edition the core should be way closer to 77 to 80 degrees, clearly they have done a bad job of repasting the gpu, I do know the ref edition uses Hitachi thermal pad, so if repasting with thermal paste , then four nylon washers are recommended to add to the mounting pressure.

I am going to use Noctua H2 paste.

I was just wondering if anyone else in the thread has repasted their reference edition rx 5700xt?
Click to expand...
Yes I did it once and saw worse temps than with the Hitachi pad, and I was using washers.

Try some washers see if it helps.
 
LtMatt said:
Yes I did it once and saw worse temps than with the Hitachi pad, and I was using washers.

Try some washers see if it helps.
Click to expand...

Thank you for your reply.

Few extra questions.

How much worse where the temps, where they as bad as 90 degrees?

Did you try with and without the washers?

Which thickness of washers did you use?
 
justwondering said:
Thank you for your reply.

Few extra questions.

How much worse where the temps, where they as bad as 90 degrees?

Did you try with and without the washers?

Which thickness of washers did you use?
Click to expand...
Is that 90c on the Edge or Junction temperature?

If it's the latter that's actually good, if it's the former then what is the Junction temperature?

I think i used plastic washers from the auction place, 0.8MM thick i think.

They helped a bit, ultimately though i wish i had just kept using the stock thermal pad.
 
LtMatt said:
Is that 90c on the Edge or Junction temperature?

If it's the latter that's actually good, if it's the former then what is the Junction temperature?

On the junction I believe it is around 110 when testing underload with Heaven benchmark on a loop... 90 is the one that is shown typically as the gpu temp on Afterburner.

I think i used plastic washers from the auction place, 0.8MM thick i think.

They helped a bit, ultimately though i wish i had just kept using the stock thermal pad.
Click to expand...
LtMatt said:
Is that 90c on the Edge or Junction temperature?

If it's the latter that's actually good, if it's the former then what is the Junction temperature?

I think i used plastic washers from the auction place, 0.8MM thick i think.

They helped a bit, ultimately though i wish i had just kept using the stock thermal pad.
Click to expand...

On the junction I believe it is around 110 when testing underload with Heaven benchmark on a loop... 90 is the one that is shown typically as the gpu temp on Afterburner.

The card functions well enough, but my temps are high compared with the big YouTubers reviews of the ref cards.

What temps did you see after pasting?
 
justwondering said:
On the junction I believe it is around 110 when testing underload with Heaven benchmark on a loop... 90 is the one that is shown typically as the gpu temp on Afterburner.

The card functions well enough, but my temps are high compared with the big YouTubers reviews of the ref cards.

What temps did you see after pasting?
Click to expand...
110c is throttling temperature. Sounds to me like the user changed the paste and it's affected the temperatures.

Entirely possible and happened to me with one 5700 XT i used as an experiment.

With the reference 5700 XT and the reference 6000 series unless you plan to go full water cooling, it's honestly better to leave the stock thermal pad in place. It's there for a reason, it does a good job.

Sounds like you will have to take it apart to try and improve things, or return it.

I can't remember the exact temps i saw, but i know they were worse.
 
Hi everyone, I've made good use of this thread in learning how to do the below, however I think I've messed it up. I've pasted this from another forum but I would appreciate some advice from fellow owners and so have posted here as well (apologies to mods if this violates any rules - I will gladly change it to comply but I think it's Ok):


I've been seeing VRAM temps on my Red Devil 5700xt in the 90-92c range under load. I read that replacing the thermal pads on the VRAM could help operating temperatures reduce by a few degrees.

After watching a couple of thermal pad replacement videos and being unable to find any clear indication of the stock termal pad sizes I settled for the 1.5mm thermal grizzly minus 8 pads. I didn't want to go for 2mm so as to avoid putting any undue stress on the components when tightening the heatsink on.

Following replacement of the pads my VRM temp appears to have almost HALVED(Wow?!) from 44c to 22c. I used mx-4 thermal past on the gpu die and am seeing a slight increase of 3 degrees or so, but still no more than 60-70c.

My worry is that I'm seeing very slightly worse VRAM peaks at 94c which I can attribute to one of the following issues:

-Thermal conductivity of stock pads is the same as that of the replacements - solution would be to get more conductive pads such as Gelid or Fujipoly

-Thickness of replacement pads was not great enough, and even if conductivity is better than stock, mounting pressure is worse due to thickness. - solution would be to buy the same or more conductive pads, only thicker.

Does anyone have experience of replacing the pads, and if so can you recommend the proper thickness and whether for example a Gelid GP extreme pad with 11W/mK rather than thermal grizzly's 8(?) W/mk would be a better idea?

BONUS - I've seen some deep forum posts around the internet suggesting using a very fine application of thermal paste between pads and VRAM - any opinions on this? Histoical opinion looks to be that it's a massive no although I've also found some manufacturers and forum posts that apparently do it/recommend it as a means of reducing air gaps between pads and chips.

tl:dr - should the red devil 5700xt thermal pads be 1.5mm or 2.0mm, and is 8w/mk an improvement on the stock pads or is greater thermal conductivity required to see a difference?

Thank you in advance for any replies.
 
You must log in or register to reply here.
Back
Top Bottom