1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

The Radeon VII Owners Thread

Discussion in 'Graphics Cards' started by Kaapstad, Feb 6, 2019.

  1. Chud_Jenny

    Gangster

    Joined: Aug 5, 2017

    Posts: 230

    Location: Cornwall

    Looks like windows update bet you to the driver install!
     
  2. Illuminist

    Mobster

    Joined: Sep 26, 2017

    Posts: 2,778

    Location: In the Masonic Temple

    Okay so I restart the pc, installed the drivers again and this time is asked me if i wanted a clean install, everything worked.
    Now in adrenaline i have 1 mv more on my core lol
    I haven't seen a monitor glitch "yet"
    Going to run timespy stress test and see whats up
    *fingers crossed*
     
  3. majestynl

    Associate

    Joined: Sep 23, 2018

    Posts: 36

    Location: Amsterdam

    Quickly tried this morning with white bars starting min 1900mhz and max at 1975mhz. Probably need both bars on top of each ?!
    And if we enable auto mem at the end, it only effects the mem. GPU core stays at default clocks.
    Will play more tonight.

    Check out the 3rd image, he was talking about same approach ;)
    https://www.techpowerup.com/reviews/AMD/Radeon_VII/33.html

    @Vince
     
  4. Vince

    Man of Honour

    Joined: Oct 30, 2003

    Posts: 7,594

    Location: Essex

    Ahh yea take a look at that. What you need to do is click apply on the manual bits first, then do auto memory, click apply again and that sets the memory. if you push more volts on the core memory will reduce to compensate.
     
  5. Illuminist

    Mobster

    Joined: Sep 26, 2017

    Posts: 2,778

    Location: In the Masonic Temple

    Okay, so the stress test failed in timespy, and crashed, i can see the temps were hitting 115 on the junction, lowered the mv by 10mv and the stress test hung.
    this card is completely useless as what happens above happens in every game, #sosadrightnow
     
  6. Chud_Jenny

    Gangster

    Joined: Aug 5, 2017

    Posts: 230

    Location: Cornwall

    Can you run a Timespy stress test at default settings crash free?
     
  7. Illuminist

    Mobster

    Joined: Sep 26, 2017

    Posts: 2,778

    Location: In the Masonic Temple

    That was a timespy stress test at default, I touched nothing, it crashed before the end and gave me a 0 failed
    The temps were 92 core and 115 junction

    I then tried with a 10mv undervolt and it instantly crashed on the stress test

    I turned the fans to 100% just now and ran the stress test at default. It passed the test but the junction was 109.c for the majority of the run.
    So In conclusion, unless I run the card at 100% fan all the time, my games will crash.
     
    Last edited: Feb 11, 2019
  8. Chud_Jenny

    Gangster

    Joined: Aug 5, 2017

    Posts: 230

    Location: Cornwall

    Ok all makes sense to me. Clock dropped when the card hit 115C, hence the test fail with the clock drop.
    Any idea what RPM the fan was running during the test?
    I'd try a 100% fan and retest, I bet it will pass due to the lower temps.
     
  9. Boomstick777

    Capodecina

    Joined: Jan 2, 2012

    Posts: 11,112

    Location: UK.

    A water-cooled version with similar aesthetics would be sweet !

    Make it happen AMD xD
     
  10. Gerard

    Caporegime

    Joined: Oct 18, 2002

    Posts: 30,860

    Location: Ireland

    They would really only need to modify the cold plate of the cooler they used for the Liquid vega 64 for that to happen. 5 more watts isn't going to make a difference to that cooler.
     
  11. Boomstick777

    Capodecina

    Joined: Jan 2, 2012

    Posts: 11,112

    Location: UK.

    I would buy a liquid cooled version with similar aesthetic the day it launched. Come on AMD make it happen !

    I'll have to time share my GPU's. RTX 2080 during the week and Radeon 7 for the weekends !! xD
     
  12. Illuminist

    Mobster

    Joined: Sep 26, 2017

    Posts: 2,778

    Location: In the Masonic Temple

    2900rpm on the first test where I didn't change the fans.

    And you quoted me, I said already I did a stress test at 100% fan and it passed but it was 109.C junction temp the whole time

    I've just set up an RMA to send it back for replacement, this is a terrible card, im not putting my fan to 100% just to use it, especially as it just crashes after time anyway
     
  13. Chud_Jenny

    Gangster

    Joined: Aug 5, 2017

    Posts: 230

    Location: Cornwall

    Sounds a good idea to RMA, i wonder if the mounting pressure isn't enough as that would tally up to techpowerup ect....
     
  14. Gerard

    Caporegime

    Joined: Oct 18, 2002

    Posts: 30,860

    Location: Ireland

    They'd have to charge more for it if they ever bothered to make one, probably £50-£100 more.
     
  15. Chud_Jenny

    Gangster

    Joined: Aug 5, 2017

    Posts: 230

    Location: Cornwall

    Be interesting to see what price the EKWB block goes on sale for @ ocuk. My cards in the cupboard til then!
     
  16. Otellino

    Associate

    Joined: Feb 11, 2019

    Posts: 4

    Hey all! I'll post photos when I'm home to hop onto the Roll of Honour, but until then I'm having the strangest of issues with my Sapphire VII, with some similarities to Illuminist's issues but not exactly the same;

    I've got a dual monitor setup. Primary is an AOC AGON AG271QX which is a 1440p 144hz FreeSync display, which I'm connected to via DisplayPort. Secondary monitor is an Asus monitor whose model name escapes me. It's nothing special, just a 24inch 1080p 60hz monitor, connected via HDMI. (It's only got that and DVI).

    Now, as far as I can tell thus far, the card is gaming just fine. No stability issues, great performance in games like Fallout 76, AC: Odyssey and R6: Siege. This is without having touched Wattman as I know there's issues with that currently. My big problem is in general web browsing (I use Firefox) certain web pages will cause my AGON (And only my AGON) to start flickering with horizontal static-looking artifacts with varying regularity.

    If I move my Firefox window to my Asus display, the artifacting stops immediately. Also, if I resize the Firefox window to approx. half my screen size on my AGON, it also stops flickering... Any ideas on if this could be a hardware or software issue?

    My system specs are the following:
    MSI Gaming Pro Carbon X399
    Threadripper 1920X
    Corsair AX860(non-i) 860 Watt PSU
    32GB of quad-channel memory at 3000Mhz.

    Any help is much appreciated! :)

    EDIT: I have tried all 3 other DisplayPort ports on my GPU. I haven't tried a different cable, but it did work fine on my previous GPU so I assume the cable isn't bad?
     
  17. Colzy

    Gangster

    Joined: Jan 31, 2014

    Posts: 149

    Location: Bristol, UK

    If firefox is the only issue it's almost certainly that. Could be a plugin you have or the hardware acceleration in it isn't performing correctly with the Radeon. You could try turning that off in the settings.
     
  18. Chud_Jenny

    Gangster

    Joined: Aug 5, 2017

    Posts: 230

    Location: Cornwall

    Best thing to do is try edge browser as even chrome isn't good at the moment.
     
  19. Otellino

    Associate

    Joined: Feb 11, 2019

    Posts: 4

    Sure, when I'm home I can give a different browser a go and report back. It's just the strangest thing - it's only on very specific websites (eg Tom's Hardware). It's the inconsistency that's confusing me as to the root of the problem, haha. Thanks!
     
  20. Scougar

    Capodecina

    Joined: Jan 30, 2007

    Posts: 12,086

    Location: PA, USA

    I think AMD did mention there were some 2nd monitor bugs going on at the moment. I know one bug is having a second connected monitor but it not being turned on caused issues for example. It's all driver related.