The Radeon RX 6800 (XT) Owners Thread.

rallyslag · 29 Jan 2021 at 08:24

LtMatt said:
Expected behaviour, it’s due to the vertical blank timing on your monitor that requires a maximum memory clock. You might find that at the desktop dropping the refresh to 120 may allow a lower clock due to the timing changing to fall within the sweet zone.

Expected behaviour. For a bug! There is no reason why my memory needs run at 2000mhz at desktop. The card draws 50 rather than 10 watts making it idle at 40 rather than 30 deg. This in turn makes my southbridge temps over 50 deg rather than 45 deg . Ok the southbridge high temps is As rocks fault,but nether the less it is still happening. And on a card i paid over £700 for. This can be in no way good for the longevity of the card. Its a shame as apart from this(big issue for me) it's a cracking card.

LtMatt · 29 Jan 2021 at 08:42

rallyslag said:
Expected behaviour. For a bug! There is no reason why my memory needs run at 2000mhz at desktop. The card draws 50 rather than 10 watts making it idle at 40 rather than 30 deg. This in turn makes my southbridge temps over 50 deg rather than 45 deg . Ok the southbridge high temps is As rocks fault,but nether the less it is still happening. And on a card i paid over £700 for. This can be in no way good for the longevity of the card. Its a shame as apart from this(big issue for me) it's a cracking card.

No it's not a bug and does not impact the longevity of the GPU in anyway.

You can try tweaking the timing values of your monitor using CRU which will adjust the vertical blank timing and may allow for a lower memory clock. Have a look at this thread.

Johnny Silverhand · 29 Jan 2021 at 08:54

rallyslag said:
I have a Sapphire nitro+ RX6800 and a 1440 @144hz monitor. The vram is running at full speed all the time but if i restrict my monitor to 60hz it steps all the way down to idle. Has anybody with a rx6800 and same spec monitor noticed the same behaviour?

This isn't spercific to AMD GPU's I don't think...? Maybe this will help.

Kaapstad · 29 Jan 2021 at 09:09

roger_37 said:
https://imgur.com/a/0xkVsnz

Sorry I can not see your image.

rallyslag · 29 Jan 2021 at 09:14

Johnny Silverhand said:
Perhaps this could help? https://i.reddit.com/r/Amd/comments/dr5bxt/i_solved_my_max_memory_clocks_at_144hz_problem/

Thanks. I'll have a play with CRU later and see if i can get something working . It does seem weird this issue has been about a while and not addressed . But if AMD consider it "expected behaviour" their is nothing they are willing or able to do about it. Reading the link you supplied though does point towards it being a monitor issue more than a gpu issue though. Strange it never happened with my Vega 64. Probably HBM memory i am guessing?

roger_37 · 29 Jan 2021 at 09:14

Kaapstad said:
Sorry I can not see your image.

im using imgur,dont know why you cant see image?

roger_37 · 29 Jan 2021 at 09:59

https://hosting.photobucket.com/ima...jpg?width=450&height=278&fit=bounds&crop=fill
hope this is better.

henson0115 · 29 Jan 2021 at 11:20

LtMatt said:
Okay cheers mate. Need to see if anyone else sees it or just local to your system.

can confirm its gen 4 pci causing this, cards has been solid all day yesterday after swapping to gen 3 both idle and in game maxed out, stress tests you name it. also had rdp running too when it was idle and its been solid. no crashes thus far today either and there is also multiple reports of gen 4 giving people grief on gigabyte x570 boards on oc.net. will stick with gen 3 pci for now. cba with faffing with vddp voltages..

LtMatt · 29 Jan 2021 at 11:36

henson0115 said:
can confirm its gen 4 pci causing this, cards has been solid all day yesterday after swapping to gen 3 both idle and in game maxed out, stress tests you name it. also had rdp running too when it was idle and its been solid. no crashes thus far today either and there is also multiple reports of gen 4 giving people grief on gigabyte x570 boards on oc.net. will stick with gen 3 pci for now. cba with faffing with vddp voltages..

Yeah seen a few reports of issues with PCI-E 4 and Gigabyte boards. I don't see that issue personally using an Asus Crosshair VIII Hero, and before that a Asus Prime Pro.

EDIT - Random suggestion, try bumping Soutbridge voltage to 1.025v-1.050v see if it helps. Stock is 1.000v.

Smoogels · 29 Jan 2021 at 11:54

henson0115 said:
anyone else having to swap from pci gen 4 to 3 to stop random reboots on a 6800/xt? cpu and memory are bone stock, only change is the gpu. happens roughly once a day and whilst in rdp of all things. seems to be in lower power modes? temps are amazing so it aint that. separate 8 pin cables for the psu and its made by seasonic.

I've been having random reboots ever since installing the RX 6800 Nitro however I'm on a X470 crosshair which is only PCIe 3. Yesterday it rebooted four times at idle.
I have it paired with a 5900X which ran perfectly with my old Nvidia card and at load theres no problems at all. The most frustrating thing is just how random it is. Just when you think it's solved, bam, slaps you with a reboot. Might be a slightly different problem since PCIe 3.0 resolved it in your case but had to post since I have similar issues with low power it seems.

LtMatt · 29 Jan 2021 at 12:11

Smoogels said:
I've been having random reboots ever since installing the RX 6800 Nitro however I'm on a X470 crosshair which is only PCIe 3. Yesterday it rebooted four times at idle.
I have it paired with a 5900X which ran perfectly with my old Nvidia card and at load theres no problems at all. The most frustrating thing is just how random it is. Just when you think it's solved, bam, slaps you with a reboot. Might be a slightly different problem since PCIe 3.0 resolved it in your case but had to post since I have similar issues with low power it seems.

Have you enabled Curve Optimizer? Can you also try entering the BIOS > AMD CBS Menu and look for an option called Power Supply Idle Control set this to Typical.

martian1 · 29 Jan 2021 at 12:34

Hoping any XFX RX6800 XT MERC319 owners can help or advise. I have a friend who managed to get the card and he's having issues which looks like his motherboard as it was acting up prior him buying the card. He has the gigabyte x570 master i believe anyway when things were working he managed to play games and run timespy etc ok but he is unsure if this card behavior is normal..basically prior the game or benchmark loading his stats app (afterburner) states the Gpu is running at zero mhz ! Not sure if this is some sort off power saving feature of the card and is normal. His psu is easily capable as its a 1000w corsair platinum.
Any of you guys getting similar who own the card ?

Thanks

henson0115 · 29 Jan 2021 at 14:12

LtMatt said:
Yeah seen a few reports of issues with PCI-E 4 and Gigabyte boards. I don't see that issue personally using an Asus Crosshair VIII Hero, and before that a Asus Prime Pro.

EDIT - Random suggestion, try bumping Soutbridge voltage to 1.025v-1.050v see if it helps. Stock is 1.000v.

will give it a shot ta.

edit fyi: just had a random reboot with rdp running and gen3.

henson0115 · 29 Jan 2021 at 14:13

Smoogels said:
I've been having random reboots ever since installing the RX 6800 Nitro however I'm on a X470 crosshair which is only PCIe 3. Yesterday it rebooted four times at idle.
I have it paired with a 5900X which ran perfectly with my old Nvidia card and at load theres no problems at all. The most frustrating thing is just how random it is. Just when you think it's solved, bam, slaps you with a reboot. Might be a slightly different problem since PCIe 3.0 resolved it in your case but had to post since I have similar issues with low power it seems.

actually its not pci3 for me, its just bloody happened again...

almost like theres not enough volts going to it in low power states or something.

LtMatt · 29 Jan 2021 at 14:14

henson0115 said:
actually its not pci3 for me, its just bloody happened again...

Hmm, the plot thickens.

Can you try updating to the latest BIOS (if not on it already) and putting the BIOS optimized default settings on? See if it happens then.

henson0115 · 29 Jan 2021 at 14:17

LtMatt said:
Hmm, the plot thickens.

Can you try updating to the latest BIOS (if not on it already) and putting the BIOS optimized default settings on? See if it happens then.

already updated from f32 to f33a yesterday to try and resolve the issue. i will try the optimized defaults shortly. although pretty much eevrythign is stock in my case otehr than the gen 3 and normal for the psu current load.

random_matt · 29 Jan 2021 at 14:28

What is the RRP? Just saw the main competitor selling the taichi for £999.99.

henson0115 · 29 Jan 2021 at 14:38

LtMatt said:
Hmm, the plot thickens.

Can you try updating to the latest BIOS (if not on it already) and putting the BIOS optimized default settings on? See if it happens then.

spotted that when i changed to gen 3 that i had reverted the current supply (it was set to low instead of normal), testing with normal again instead of low now. although not sure why it wouldn't be a problem on the Vega too, seems very odd.
even worse i don't get any event logs, but i will try normal first, then try optimized.
might even try an older bios pre re size bar changes, im curious if that will help too.

edit: i have also noticed that after the reboot the amd software gives me a message to say that the graphics settings have been reset due to a failure. other than a custom curve, its all bone stock.

Kaapstad · 29 Jan 2021 at 15:53

roger_37 said:
https://hosting.photobucket.com/ima...jpg?width=450&height=278&fit=bounds&crop=fill
hope this is better.

Which card is it please.

stu.artd · 29 Jan 2021 at 15:56

random_matt said:
What is the RRP? Just saw the main competitor selling the taichi for £999.99.

Due to increased costs throughout the GPU supply chain, the RRP seems to be a thing of the past.

I looked at that card. Although I really do need to move on from my old 7970 I just can’t justify spending that much on a 6800 xt or 3080. Or £700+ on a 6800 or 3070.