Made a reasonable difference on my old Sapphire 270X's, and on my old Asus 390X Strix the difference was substantial.
Its not a difficult job to do so worst case is I've spent an hour or two doing it and the difference is minimal.
-
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 RX Vega 56 Owners Thread
- Thread starter Kaapstad
- Start date
Made a reasonable difference on my old Sapphire 270X's, and on my old Asus 390X Strix the difference was substantial.
Its not a difficult job to do so worst case is I've spent an hour or two doing it and the difference is minimal.
Example: https://www.reddit.com/r/Amd/comments/cro572/replacing_the_thermal_pads_and_paste_on_my_vega/
lower temps and lower fan speed.
lower temps and lower fan speed.
Might have been worth it on other cards? But anyhow up to you, gluck with it and hopefully you see some benefit. 
https://www.techpowerup.com/forums/...anging-thermal-paste-and-thermal-pads.246646/ maybe worth a read also.
Especially this post.
https://www.techpowerup.com/forums/...al-paste-and-thermal-pads.246646/post-3901491
https://www.techpowerup.com/forums/...anging-thermal-paste-and-thermal-pads.246646/ maybe worth a read also.
Especially this post.
https://www.techpowerup.com/forums/...al-paste-and-thermal-pads.246646/post-3901491
I got it installed and working last night.
I haven't pushed it to it's limit yet (it was getting late - very late 
) but it's running just slightly below your recommended speeds at the voltages you recommended, with RAM at 950 (it's Micron, btw). Thanks again for the help.The temps are still quite low, though. I'll have to check the fan curve again and do more testing but the highest I saw during benchmarks (Timespy and Firestrike) was about 62C. There's a bit of coil whine sometimes too but it's not too bad.
Anyway, it's a cracking little card. I'm very happy with it.
Last edited:
Soldato
- Joined
- 21 Jul 2005
- Posts
- 20,701
- Location
- Officially least sunny location -Ronskistats
Soldato
- Joined
- 21 Jul 2005
- Posts
- 20,701
- Location
- Officially least sunny location -Ronskistats
I think mines Hynix, but it would not be comfortable over 950 either.
Samsung for me. I must have lucked out on the silicon lottery with mine reaching 980. TBH though, the extra 30mhz will only have small benefit. If I remember correctly, moving from 800 to 980 gained an extra 2 fps in Division 2 1440p ultra
Soldato
- Joined
- 21 Jul 2005
- Posts
- 20,701
- Location
- Officially least sunny location -Ronskistats
Might make more of a difference on memory aggressive apps. I think I settled on 920 as its not crashed on anything for months thrown at it and the temps seem comfortable too.
You're welcome. Glad it is working well. Out of interest, what res and refresh rate do you game on and what games are you playing?I got it installed and working last night.
The temps are still quite low, though. I'll have to check the fan curve again and do more testing but the highest I saw during benchmarks (Timespy and Firestrike) was about 62C. There's a bit of coil whine sometimes too but it's not too bad.
Anyway, it's a cracking little card. I'm very happy with it.
I think there's often a tendency to push the oc to the max when first getting a new component but this doesnt always give the best overall experience. I did this when I got my pulse last year trying to get a solid 60fps+ with Division 2 on 1440 ultra. I pretty much played for the first month on 1672 mhz / 980 mhz / 50% PL. The card alone would draw 260w and the fan noise was really intrusive even over headphones. Plus it would crash periodically. Just dialling it back to 1662 mhz was a significant improvement in terms of stability and noise with only an imperceptibly small reduction in fps. Now I run at 1637 and 15% PL with the fog and reflections settings dropped down a notch to compensate. I just ran a couple of Div2 benchmarks there and the difference between 1662/50% and 1637/15% was -2 fps (-3%) but the power draw was -14%, temp -5% and fan rpm -10%. For me that is an acceptable trade-off
@5t3v0 I'm at 1080p/60 at the moment but a 1440p screen is on the cards soon (not sure what refresh rate yet). BFV is probably the most demanding game I'll be playing.
This isn't all about practicality though. I've always tinkered with my hardware (fnar) whether there's any need to or not.
This isn't all about practicality though. I've always tinkered with my hardware (fnar) whether there's any need to or not.
Just ran some tests whilst I had the Div2 benchmark open and 980, 950 & 920 all came out at 62 fps on 1440 ultra. The default 800 mhz was 60 fps
Does anyone know the score with Micron HBM modules on Vega 56 cards?
I've read two conflicting opinions:
1) They're not as good as the desirable Samsung modules (therefore similar to Hynix modules).
2) They're actually Samsung modules being misreported as Micron by an old version of GPUZ. (I'm running an old version.)
I'll download the latest GPUZ when I get on the PC later and have a look.
I've read two conflicting opinions:
1) They're not as good as the desirable Samsung modules (therefore similar to Hynix modules).
2) They're actually Samsung modules being misreported as Micron by an old version of GPUZ. (I'm running an old version.)
I'll download the latest GPUZ when I get on the PC later and have a look.
Soldato
- Joined
- 21 Jul 2005
- Posts
- 20,701
- Location
- Officially least sunny location -Ronskistats
Try the power limit on 15% and see if your performance feathers but the consumption drops much with it?
Me too - I am fanatical min maxer - but I found it more fun to find the sweet spot where performance-power balance was most optimal as this was a much harder challenge than finding the highest OC. I benchmarked a number of games at loads of setting combinations and plotted lots of tables and graphs which you can see in my earlier posts in this thread. I now use a couple of OC profiles which I switch between depending on how demanding a particular game is. My target now instead of maxing fps is getting the the best quality settings while keeping within the 48-75 hz freesync range of my monitor@5t3v0 I'm at 1080p/60 at the moment but a 1440p screen is on the cards soon (not sure what refresh rate yet). BFV is probably the most demanding game I'll be playing.
This isn't all about practicality though. I've always tinkered with my hardware (fnar) whether there's any need to or not.
So on Div2 1440/ultra with 1662 gpu and 980 vram, 50% PL gives 62 fps for 243W, 65C and 2080rpm, 15% PL gives 60 fps for 207W, 64C and 1960rpm. So from 50% to 15% fps drops 3% for 16% drop in power, 3% drop in temp and 6% reduction in fan rpm.
Now that's enough min-maxing for tonight
Yes, I see. There's clearly more to tuning Vega than going for high clocks.
It's very early days for me though - it's been in my PC less than 24 hours at this point and I've only had a few hours playing with it (midnight to *cough* 4am this morning). I'm still very much at the "how fast can it go" stage.
LOL. You're absolutely right mate. It took me 3 months of experimenting and indecision before I settled on what I use now. I shouldn't discourage you from that enjoyment. Its a great card to tweak around with so enjoy and keep us posted on how you get onYes, I see. There's clearly more to tuning Vega than going for high clocks.
It's very early days for me though - it's been in my PC less than 24 hours at this point and I've only had a few hours playing with it (midnight to *cough* 4am this morning). I'm still very much at the "how fast can it go" stage.
Div 2 is easy with 4k on vega 56 imo.