Overclock.net banner
1 - 8 of 8 Posts

·
High IQ Person
Joined
·
627 Posts
Discussion Starter · #1 ·
I'm having a strange issue with my old 480 GTR BE (580 custom bios), it seems that core voltage goes over the usual stock of 1.25v max.

Grey Font Flash photography Darkness Logo

Flash photography Font Darkness Monochrome photography Monochrome


Temps are super chill so I'm not sure if it's even real, could it be just a bug or error in hwinfo?
If not, mb something wrong with the vrms? The card is in really good conditions tho.
 

·
Iconoclast
Joined
·
31,443 Posts
BIOS from a card with the same VRM PWM controller? Was droop messed with in the custom BIOS?

Could well be a reporting error, but custom firmware raises some questions.
 

·
High IQ Person
Joined
·
627 Posts
Discussion Starter · #3 · (Edited)
BIOS from a card with the same VRM PWM controller? Was droop messed with in the custom BIOS?

Could well be a reporting error, but custom firmware raises some questions.
I'm using this bios for about 3+ years now, I did some edits but is based on the XFX RX580 GTR BE, in theory exaclty the same PCB/Compenents just with a newer Polaris chip.

No voltage or any type of droop was changed, just the TDP and clocks, but the strange things is I've never had this voltage reports before, so Idk if is something I should care about or just live with it.
 

·
Iconoclast
Joined
·
31,443 Posts
I would try different firmware just to see if it changes the reported voltage behavior. If it does, it may be worth looking into further if you plan on keeping the card long enough for it to matter (years). If not, you can probably put it down to a sensor or software issue.
 

·
Registered
Joined
·
2,936 Posts
I would try different firmware just to see if it changes the reported voltage behavior. If it does, it may be worth looking into further if you plan on keeping the card long enough for it to matter (years). If not, you can probably put it down to a sensor or software issue.
I wouldnt say years. 1.4v through that card can be detrimental. Hell even my 15 year old 55nm 3870 doesnt have 1.4vcore stock lol.
 

·
Iconoclast
Joined
·
31,443 Posts
I wouldnt say years. 1.4v through that card can be detrimental. Hell even my 15 year old 55nm 3870 doesnt have 1.4vcore stock lol.
I'm under the impression that the voltage, even if real, is transient. A 1.4v set core voltage under load would probably kill the part in short order and would definitely result in noticeably high temperatures.

The card is already years old and I'm doubtful anything on the card has changed that would spontaneously result in a higher output voltage from the VRM. There are failure modes that could do this, but it's more likely some HWiNFO update that introduced some reporting bug.
 

·
Registered
Joined
·
2,936 Posts
I'm under the impression that the voltage, even if real, is transient. A 1.4v set core voltage under load would probably kill the part in short order and would definitely result in noticeably high temperatures.

The card is already years old and I'm doubtful anything on the card has changed that would spontaneously result in a higher output voltage from the VRM. There are failure modes that could do this, but it's more likely some HWiNFO update that introduced some reporting bug.
I totally agree but with bios mods 1.4 is possible from what i recall. Ovp kicks in above that though iirc
 

·
High IQ Person
Joined
·
627 Posts
Discussion Starter · #8 ·
Looks like random spikes, never steady coz avg voltage is 1.2~ and doesn't seems to affect temps, still 75c as bios target.
I might try another bios, but tbh the only change I did a few months ago is the UEFI compatibility mod (AMDGOP1.69.0.15.50), that's it.

Anyways, about what @o1dschoo1 said, yep is 100% possible to tweak voltages and also add LLC, did it in the past and testest till 1.35v, but is not worth it, kinda 10mhz every 0.05v.

I can post my current bios if anyone wants to take a look at it.
 
1 - 8 of 8 Posts
Top