Overclock.net › Forums › Graphics Cards › AMD/ATI › RX 480 overclock discussion
New Posts  All Forums:Forum Nav:

RX 480 overclock discussion - Page 14

post #131 of 433
Quote:
Originally Posted by greytoad View Post

that voltage makes sense. Only pretty good asics can run 1400 at 1150. Thanks. Nice card.

What Asics would be considered good ... i got a 470 4G nitro + the other day and went to check the asics but haven't done any OC tuning just wanted to enjoy it a bit without messing with it.
Got 92.8 or was it 89.2 oh ... smile.gif seemed quite high.

I read that different GPU generations score differently on average. What are the averages for elesmere?
Edited by slavovid - 8/22/16 at 6:10am
Home PC
(8 items)
 
  
CPUMotherboardGraphicsRAM
FX 8350  M5A99FX PRO R2.0 GTX 650 2x4GB DDR3 2400 mhz 11-13-13-35 
Hard DriveCoolingOSPower
Several Thermalright Inferno IFX-14 + 2x ML 120's + 2x... Windows 10 Cooler Master B700 v.2 
  hide details  
Reply
Home PC
(8 items)
 
  
CPUMotherboardGraphicsRAM
FX 8350  M5A99FX PRO R2.0 GTX 650 2x4GB DDR3 2400 mhz 11-13-13-35 
Hard DriveCoolingOSPower
Several Thermalright Inferno IFX-14 + 2x ML 120's + 2x... Windows 10 Cooler Master B700 v.2 
  hide details  
Reply
post #132 of 433
depends on what your planning on doing with the card, but the rx480 is limited mostly by power from the motherboard, not cooling. So a high ASIC can get a higher safe overclock even on water because voltage will be lower for same speed and voltage is limited from the motherboard. In a power unlimited cooling unlimited scenario a lower asic would be better as it could stable overclock higher. That's not a common situation even in water cooling for this card. I'm limited on water by safe(ish) power output levels from the motherboard not by stable clocks. I'm sure there's a sweet spot somewhere for the right amount, but I only have the 1 card at a 80.9 asic. At 1.233volts my card uses around 240watts in average gaming but I get spikes as high as 280watts. (Power usage measured by a kilawatt P4400. I take measure of the total system power and guestimate based on how much my system uses on stock with the video card on load vs. my overclock and assuming card on stock uses about 155 watt on stock. So I add the addition total system wattage to 155 after applying my overclock to get to 240. Not 100% accurate but probably closer than gpuz) Going higher isn't really safe for my motherboard so I'm power limited. 3dmark and Metro Last Light are higher. Heck I measured closer to 320 watt last night for a min. A higher asic might allow a higher 24/7 overclock at that voltage, but might also limit how high you could go on suicide bench runs. I'd rather have the higher 24/7.
Edited by greytoad - 8/22/16 at 6:17am
post #133 of 433
Quote:
Originally Posted by greytoad View Post

depends on what your planning on doing with the card, but the rx480 is limited mostly by power from the motherboard, not cooling. So a high ASIC can get a higher safe overclock even on water because voltage will be lower for same speed and voltage is limited from the motherboard. In a power unlimited cooling unlimited scenario a lower asic would be better as it could stable overclock higher. That's not a common situation even in water cooling for this card. I'm limited on water by safe(ish) power output levels from the motherboard not by stable clocks. I'm sure there's a sweet spot somewhere for the right amount, but I only have the 1 card at a 80.9 asic. At 1.233volts my card uses around 240watts in average gaming but I get spikes as high as 280watts. (Power usage measured by a kilawatt P4400. I take measure of the total system power and guestimate based on how much my system uses on stock with the video card on load vs. my overclock and assuming card on stock uses about 155 watt on stock. So I add the addition total system wattage to 155 after applying my overclock to get to 240. Not 100% accurate but probably closer than gpuz) Going higher isn't really safe for my motherboard so I'm power limited. 3dmark and Metro Last Light are higher. Heck I measured closer to 320 watt last night for a min. A higher asic might allow a higher 24/7 overclock at that voltage, but might also limit how high you could go on suicide bench runs. I'd rather have the higher 24/7.

Oh i am tempted to see if i can get 1400 24/7 oc now ... 225w on the 470 should be more than enough biggrin.gif
Home PC
(8 items)
 
  
CPUMotherboardGraphicsRAM
FX 8350  M5A99FX PRO R2.0 GTX 650 2x4GB DDR3 2400 mhz 11-13-13-35 
Hard DriveCoolingOSPower
Several Thermalright Inferno IFX-14 + 2x ML 120's + 2x... Windows 10 Cooler Master B700 v.2 
  hide details  
Reply
Home PC
(8 items)
 
  
CPUMotherboardGraphicsRAM
FX 8350  M5A99FX PRO R2.0 GTX 650 2x4GB DDR3 2400 mhz 11-13-13-35 
Hard DriveCoolingOSPower
Several Thermalright Inferno IFX-14 + 2x ML 120's + 2x... Windows 10 Cooler Master B700 v.2 
  hide details  
Reply
post #134 of 433
I meant power from motherboard not voltage from motherboard limited. doh. I'm not sure how the driver reads power limit. Power usage for the card seems higher than the power limit the driver sees. I think it's how it reads the vrm and doesn't take heat waste inefficiencies and vram power into account. Good luck with your oc. If you're on air hopefully it's not too loud. On the reference cooler I could do 1380 at 1150, but it was too loud to be a comfortable 24/7.
post #135 of 433
So what i managed to dig around is that the ASICS quality is somewhat a random number that's used for binning so if 1 generation has numbers between 60-80 and 80 being the best one another generation can have numbers between 80-100 and 100 being the best one

Yet in terms of actual overclocking capability the 80 in 1-st case is the "great" one where the 80 in the 2-nd case is the "bad" one.
So in order to determine a specific GPU asics quality we need to know the averages for that GPU generation and compare it to them. Correct?
Home PC
(8 items)
 
  
CPUMotherboardGraphicsRAM
FX 8350  M5A99FX PRO R2.0 GTX 650 2x4GB DDR3 2400 mhz 11-13-13-35 
Hard DriveCoolingOSPower
Several Thermalright Inferno IFX-14 + 2x ML 120's + 2x... Windows 10 Cooler Master B700 v.2 
  hide details  
Reply
Home PC
(8 items)
 
  
CPUMotherboardGraphicsRAM
FX 8350  M5A99FX PRO R2.0 GTX 650 2x4GB DDR3 2400 mhz 11-13-13-35 
Hard DriveCoolingOSPower
Several Thermalright Inferno IFX-14 + 2x ML 120's + 2x... Windows 10 Cooler Master B700 v.2 
  hide details  
Reply
post #136 of 433
Here's what I'm going by on asic:

https://www.reddit.com/r/pcgaming/comments/4rzw7j/the_stilt_explains_asic_quality_on_amd_cards/


So on air you may be better off with the lower ASIC.
post #137 of 433
I'm sure there's a sweet spot for different cooling solutions. I'm not cooling limited so a higher asic might work well for me but I'm not sure about higher total power draw due to higher heat waste. I didn't take that into account.
post #138 of 433
Quote:
Originally Posted by greytoad View Post

Here's what I'm going by on asic:

https://www.reddit.com/r/pcgaming/comments/4rzw7j/the_stilt_explains_asic_quality_on_amd_cards/


So on air you may be better off with the lower ASIC.


Oh ... so according to this i won't be overclocking this card at all .... bad luck Brian. biggrin.gif
Home PC
(8 items)
 
  
CPUMotherboardGraphicsRAM
FX 8350  M5A99FX PRO R2.0 GTX 650 2x4GB DDR3 2400 mhz 11-13-13-35 
Hard DriveCoolingOSPower
Several Thermalright Inferno IFX-14 + 2x ML 120's + 2x... Windows 10 Cooler Master B700 v.2 
  hide details  
Reply
Home PC
(8 items)
 
  
CPUMotherboardGraphicsRAM
FX 8350  M5A99FX PRO R2.0 GTX 650 2x4GB DDR3 2400 mhz 11-13-13-35 
Hard DriveCoolingOSPower
Several Thermalright Inferno IFX-14 + 2x ML 120's + 2x... Windows 10 Cooler Master B700 v.2 
  hide details  
Reply
post #139 of 433
I don't think that's accurate. The additional power draw from leakage seems logarithmic. So the sweet spot moves based on cooling. You'll probably be good. thumb.gif
edit: Perhaps, finfet is measured differently I don't know but I know the higher asic cards seem to be hitting a higher stable overclock at a lower voltage, but maybe the power usage is higher for the same voltage. If that's the case it's a wash depending on where the power use progression falls on different leakage levels. I was a little off on my statement above I guess. edit2: like the opposite of correct I guess. lol
Edited by greytoad - 8/22/16 at 8:10am
post #140 of 433
Quote:
Originally Posted by greytoad View Post

I don't think that's accurate. The additional power draw from leakage seems logarithmic. So the sweet spot moves based on cooling. You'll probably be good. thumb.gif
edit: Perhaps, finfet is measured differently I don't know but I know the higher asic cards seem to be hitting a higher stable overclock at a lower voltage, but maybe the power usage is higher for the same voltage. If that's the case it's a wash depending on where the power use progression falls on different leakage levels. I was a little off on my statement above I guess. edit2: like the opposite of correct I guess. lol

Oh good thing i was sleepy last night and didn't remember properly. Just got home and checked 82.8% asics quality seems average would be ok for a small overclock without rising the temperatures or sound too much ... should i just increase the Mhz without touching the volts to 1300 instead of 1260 and test what happens ... + mabe a little extra power limit?
Home PC
(8 items)
 
  
CPUMotherboardGraphicsRAM
FX 8350  M5A99FX PRO R2.0 GTX 650 2x4GB DDR3 2400 mhz 11-13-13-35 
Hard DriveCoolingOSPower
Several Thermalright Inferno IFX-14 + 2x ML 120's + 2x... Windows 10 Cooler Master B700 v.2 
  hide details  
Reply
Home PC
(8 items)
 
  
CPUMotherboardGraphicsRAM
FX 8350  M5A99FX PRO R2.0 GTX 650 2x4GB DDR3 2400 mhz 11-13-13-35 
Hard DriveCoolingOSPower
Several Thermalright Inferno IFX-14 + 2x ML 120's + 2x... Windows 10 Cooler Master B700 v.2 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD/ATI
Overclock.net › Forums › Graphics Cards › AMD/ATI › RX 480 overclock discussion