-
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.
- Thread starter Kaapstad
- Start date
Caporegime
Interesting, that's one of the lower stock clocks I've seen out of the box. I guess that's probably why someone returned it allowing you to pick up a bargain.Oh cool. Here they are:
Let me know what's the lowest voltage you can run at 1950Mhz without instability.
Interesting, that's one of the lower stock clocks I've seen out of the box. I guess that's probably why someone returned it allowing you to pick up a bargain.
Let me know what's the lowest voltage you can run at 1950Mhz without instability.
Maybe so. I've set the clock to 1950MHz and 1025mV. I've run Heaven a few times and it settles at ~1903MHz and doesn't crash.
Haven't tried any voltages between 1000mV-1025mV though.
Caporegime
Heaven is a poor test for stability. Play any demanding game that pegs GPU utilisation at 99% solid for 30-60 minutes for a true test of stability.
Ah okay. I haven't fiddled with this stuff for years as you may be able to tell.
Apparently Dirt Rally lets you loop the benchmark so I'm downloading it now. I don't have time at the mo to play for that long.
Caporegime
That will be better, but for a true test go for the above - when you have time.
Cheers its an awesome card!
Had one problem its way longer than my vega 56 so i had to drill new holes to mount my aio (was in a brand new case as well
)
Cheers its an awesome card!
Had one problem its way longer than my vega 56 so i had to drill new holes to mount my aio (was in a brand new case as well
)
Hey,
Did it worth to move from Vega56 to 5700xt, gaming wise?
That will be better, but for a true test go for the above - when you have time.
I don't know if it's the Radeon drivers or something else, but Dirt Rally keeps crashing even after setting everything to default in the Radeon software :/
EDIT: The looped benchmark has been running longer after I configured Dirt Rally in fullscreen mode. Previously I had it in windowed mode so that I could see the Radeon software at the same time. I'm now using the ALT+R overlay. Still running at default settings to see if the card works properly.
EDIT #2: Dirt Rally just crashed again with the card running at default settings. I'll try the 20.2.2 drivers.
EDIT #3: Seems like the looped benchmark is running better with the 20.2.2 drivers with the card at default settings. Oddly enough, the defaults you asked about earlier are higher in this version:
Last edited:
Caporegime
Those clocks look more like it, not a bad sample after all.I don't know if it's the Radeon drivers or something else, but Dirt Rally keeps crashing even after setting everything to default in the Radeon software :/
EDIT: The looped benchmark has been running longer after I configured Dirt Rally in fullscreen mode. Previously I had it in windowed mode so that I could see the Radeon software at the same time. I'm now using the ALT+R overlay. Still running at default settings to see if the card works properly.
EDIT #2: Dirt Rally just crashed again with the card running at default settings. I'll try the 20.2.2 drivers.
EDIT #3: Seems like the looped benchmark is running better with the 20.2.2 drivers with the card at default settings. Oddly enough, the defaults you asked about earlier are higher in this version:
Those clocks look more like it, not a bad sample after all.
Hopefully not
I used your settings and had to settle on 1030mV for 1950Mhz (1903-1908MHz actual). It's been looping Dirt Rally's benchmark for hours without a hitch, and it's nice and quiet.
Not sure what's wrong with the 20.4.1 drivers, but my system clearly doesn't like them.
Caporegime
Could be something to do with the different click speeds or voltages that were displayed.Hopefully not
I used your settings and had to settle on 1030mV for 1950Mhz (1903-1908MHz actual). It's been looping Dirt Rally's benchmark for hours without a hitch, and it's nice and quiet.
Not sure what's wrong with the 20.4.1 drivers, but my system clearly doesn't like them.
Yeah. I'll test the next version when they're released and see if they're any better.
I'll stop boring you now, but it runs just fine if I set it to 2050MHz and 1135mV (2003-2008MHz actual clock speed). I do have to bump the fan speed up to 42% @ 90c though because the junction temp gets rather high. It gets me an extra 5FPS in the Shadow of the Tomb Raider benchmark, but I'll probably use the 1950MHz overclock most of the time.
So... the washer mod works quite well.
Stock clocks (2,058MHz) undervolted to 1,100mV (down from 1,200) is giving me core temps in the 80s with ~2100RPM fan speed.
But there's a caveat. Junction temps are still in the 90s, running at a 13 or 14 degree delta to core temperatures. I saw a suggestion elsewhere that this could be caused by uneven mounting pressure? If so, I'll redo the mod at some point. It'll do for now.
Day-to-day, I'll probably stick with my "cool and quiet" settings; 1,750MHz & 875mV, resulting in low power draw (<120W during stress testing), low temps (happily under 80C), and slow fan speeds (<1,550). It's still more than enough power to run the games I regularly play at max, or near max, settings (only got a 60Hz monitor, so 60fps is enough).
Stock clocks (2,058MHz) undervolted to 1,100mV (down from 1,200) is giving me core temps in the 80s with ~2100RPM fan speed.
But there's a caveat. Junction temps are still in the 90s, running at a 13 or 14 degree delta to core temperatures. I saw a suggestion elsewhere that this could be caused by uneven mounting pressure? If so, I'll redo the mod at some point. It'll do for now.
Day-to-day, I'll probably stick with my "cool and quiet" settings; 1,750MHz & 875mV, resulting in low power draw (<120W during stress testing), low temps (happily under 80C), and slow fan speeds (<1,550). It's still more than enough power to run the games I regularly play at max, or near max, settings (only got a 60Hz monitor, so 60fps is enough).
Last edited:
Caporegime
That's okay pal i could talk about clocks and voltages all day long.
Nice.
What model are you using?
Caporegime
I'm happy with 2050/1070, core stays around 50 and junction maxes at mid 80s. This is under water.
Nice result! Would you mind linking to the washers you used please? I may attempt it at some point.
MSI Reference.
Amazed at just how quiet I've got it running. I've spent the last few weeks debating low hassle, cheap ways of fixing the noise (G12, Accelero, or swapping to an AIB or 2070S). So suddenly finding out I can do it for the cost of a tube of Kryonaught was a revelation
They were in a box of PC bits. I've got no idea where they came from, sorry.
I do, however, have four more. I'll double check that they aren't important for something I've got stored. If they aren't, you can have them for the cost of a stamp.
That's very kind. Thanks very much!