Overclock.net › Forums › Graphics Cards › NVIDIA › How To Flash A Different BIOS On Your 1080 Ti.
New Posts  All Forums:Forum Nav:

How To Flash A Different BIOS On Your 1080 Ti. - Page 23

post #221 of 333
.
post #222 of 333
Quote:
Originally Posted by Hulio225 View Post

I think there are bins below that whatever, but load temperatures don't affect my overclock... If I set 2100 MHz I won't drop to a lower clock because temperarures that's what I tried to say

Edit:I already saw your post in that thread. If you are getting less power limited and getting lower performance and scores it's the same I found out... Less power limit for the trade off of performance... Can be an advantage in games but is an disadvantage in benchmark scores

Actually edit. The scores stayed the same. They began to increase as the card regained scaling capacity at above 2150 The advantage of the Palit bios can only felt as Slim pointed out in instances where the pwr reports are more frequent. Probably not in usual gaming. The only point of contention is that I don't feel the BIOS scores better or worse than another. To me the situation and scores are identical. Shoot look at my post again. Only when pwr is being seen more often that the Palit begins to show gains but otherwise my scores are nearly identical.
Edited by KraxKill - 5/1/17 at 11:31am
post #223 of 333
2151 @ 1.093v +550 mem
http://www.3dmark.com/fs/12477286


Running higher mem here would make my score lower and lower still as ram speeds increased and pwr seen more and more often.

2151 @ 1.093v +600mem - Palit BIOS (scaling returns)

http://www.3dmark.com/fs/12499663

On the Palit, my scaling returns and the card begins to scale again with higher mem speeds.

I did the below...to confirm how pwr instances affect my scores. I clocked my card even higher to 2164 at the same 1.093v where the card was now drawing even more power. And as you can see, I pull and pull on the mem clock and the scores increase as pwr instances decrease. This is on the stock FE bios.



You have to actually have to need the power the Palit offers otherwise there is no need to go to it especially if you are seeing lower performance with it. This again was not the case for me, but if it is for you don't do it. My situation is a little different as I'm the only one I've seen drawing that much wattage. As you can see by my post, 2100-2151 is where the pwr limit begins to be easily observable. A few places at 2100 and more and more places as clocks increase along with instances of pwr.

For me to fathom how the Palit can score lower while drawing more power provided all other variables are the same (at the very least bins voltage temperature) is beyond my understanding of physics and electrical engineering.
Edited by KraxKill - 5/1/17 at 11:32am
post #224 of 333
Quote:
Originally Posted by KraxKill View Post


...

On the Palit, my scaling returns and the card begins to scale again with higher mem speeds.

I did this...to confirm how pwr instances affect my scores. I clocked my card even higher to 2164 at the same 1.093v where the card was now drawing even more power. And as you can see, I pull and pull on the mem clock and the scores increase as pwr instances decrease. This is on the stock FE bios.



You have to actually have to need the power the Palit offers otherwise there is no need to go to it especially if you are seeing lower performance with it. This again was not the case for me, but if it is for you don't do it. My situation is a little different as I'm the only one I've seen drawing that much wattage. As you can see by my post, 2100-2151 is where the pwr limit begins to be easily observable.

For me to fathom how the can score lower while drawing more power provided all other variables are the same is beyond my understanding of physics and electrical engineering.

http://www.3dmark.com/fs/12490144 <- your link with the 22341 score (Graphics Score 33 410)
http://www.3dmark.com/fs/12489719 <- your link with the 22697 score (Graphics Score 32 332)

In your lower full run score you actually have higher GPU score, so everything you said is kinda wrong since your cpu and combined scores simply yielded the higher score in total while having lower gpu score...
thats why my test, testing simply one game test 2 in time spy comparing actual fps to each other is the way better method to determine stuff cuz i just have "one" variable...
    
CPUMotherboardGraphicsRAM
i7 7700k Asus Maximus IX Apex 1080 Ti Founders Edition G.Skill DDR4 B-Die 
Hard DriveHard DriveCoolingOS
Samsung Evo 960 Nvme Western Digital RED Custom Loop Windows 10 
MonitorKeyboardPowerCase
Samsung 4K Logitech G15 Fractal Design Tesla R2 650W Enthoo Primo 
MouseAudioAudioAudio
Logitech G500 Asus Xonar STX Beyerdynamic DT 1990 PRO Headphones Blue Snowball Mic (Alu Edition) 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
i7 7700k Asus Maximus IX Apex 1080 Ti Founders Edition G.Skill DDR4 B-Die 
Hard DriveHard DriveCoolingOS
Samsung Evo 960 Nvme Western Digital RED Custom Loop Windows 10 
MonitorKeyboardPowerCase
Samsung 4K Logitech G15 Fractal Design Tesla R2 650W Enthoo Primo 
MouseAudioAudioAudio
Logitech G500 Asus Xonar STX Beyerdynamic DT 1990 PRO Headphones Blue Snowball Mic (Alu Edition) 
  hide details  
Reply
post #225 of 333
Quote:
Originally Posted by Hulio225 View Post

http://www.3dmark.com/fs/12490144 <- your link with the 22341 score (Graphics Score 33 410)
http://www.3dmark.com/fs/12489719 <- your link with the 22697 score (Graphics Score 32 332)

In your lower full run score you actually have higher GPU score, so everything you said is kinda wrong since your cpu and combined scores simply yielded the higher score in total while having lower gpu score...
thats why my test, testing simply one game test 2 in time spy comparing actual fps to each other is the way better method to determine stuff cuz i just have "one" variable...

Sure what you are seeing are the highest score out of three runs there are lower scores of each test. I try not to post the outliers. But there is no way I will score anywhere near 33K GPU score in FS on the FE Bios at 2164 +600. My overall scores are somewhere in the 22,300 range there vs near 23K on the Palit. I will reflash the stock bios and re run the test using as you suggest the test 2 in time spy bios vs bios bin for bin and see how I fare.
post #226 of 333
Quote:
Originally Posted by KraxKill View Post

Sure what you are seeing are the highest score out of three runs there are lower scores of each test. I try not to post the outliers. But there is no way I will score anywhere near 33K GPU score in FS on the FE Bios at 2164 +600. My overall scores are somewhere in the 22,300 range there vs near 23K on the Palit. I will reflash the stock bios and re run the test using as you suggest the test 2 in time spy bios vs bios bin for bin and see how I fare.

yeah do that and post your honest results, cuz (i did over for sure over 1000 time spy / fire strike full runs in total) full runs vary to much especially physics score...
in time spy my physics score differ from 6850 to 7060ish wich can give you results from 10300 -10450 at the end what is alot if you compete with other people etc.

Edit: If those are the highest scores out of three runs, your conclusion was even more wrong in my books since you are getting significantly higher gpu scores with 2163core/+400vram than 2163/+0

i guess you will conclude the same like me: palit bios doesn't give you more than a bit less power limit and worse scores.
ill definetly try some more bios and do some datasheets to see which bios is the best for benching in terms of scores
Edited by Hulio225 - 5/1/17 at 12:24pm
    
CPUMotherboardGraphicsRAM
i7 7700k Asus Maximus IX Apex 1080 Ti Founders Edition G.Skill DDR4 B-Die 
Hard DriveHard DriveCoolingOS
Samsung Evo 960 Nvme Western Digital RED Custom Loop Windows 10 
MonitorKeyboardPowerCase
Samsung 4K Logitech G15 Fractal Design Tesla R2 650W Enthoo Primo 
MouseAudioAudioAudio
Logitech G500 Asus Xonar STX Beyerdynamic DT 1990 PRO Headphones Blue Snowball Mic (Alu Edition) 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
i7 7700k Asus Maximus IX Apex 1080 Ti Founders Edition G.Skill DDR4 B-Die 
Hard DriveHard DriveCoolingOS
Samsung Evo 960 Nvme Western Digital RED Custom Loop Windows 10 
MonitorKeyboardPowerCase
Samsung 4K Logitech G15 Fractal Design Tesla R2 650W Enthoo Primo 
MouseAudioAudioAudio
Logitech G500 Asus Xonar STX Beyerdynamic DT 1990 PRO Headphones Blue Snowball Mic (Alu Edition) 
  hide details  
Reply
post #227 of 333
.
Edited by SlimJ87D - 5/1/17 at 12:44pm
post #228 of 333
Quote:
Originally Posted by Hulio225 View Post

yeah do that and post your honest results, cuz (i did over for sure over 1000 time spy / fire strike full runs in total) full runs vary to much especially physics score...
in time spy my physics score differ from 6850 to 7060ish wich can give you results from 10300 -10450 at the end what is alot if you compete with other people etc.

Edit: If those are the highest scores out of three runs, your conclusion was even more wrong in my books since you are getting significantly higher gpu scores with 2163core/+400vram than 2163/+0

i guess you will conclude the same like me: palit bios doesn't give you more than a bit less power limit and worse scores.
ill definetly try some more bios and do some datasheets to see which bios is the best for benching in terms of scores

So I'm averaging 64.6fps on the FE over 5 runs and 64.3 on the Palit over the same in Time Spy test 2. This is at 2151 @ 1.093v on both. So the FE does seem to score 1/3rd FPS higher at least in this test but it is riddled with pwr which IMO transfers onto the screen. Do you know of a good way to test frame time? Because I certainly prefer the green free Palit GPUz graphs.
post #229 of 333
Quote:
Originally Posted by KraxKill View Post

So I'm averaging 64.6fps on the FE over 5 runs and 64.3 on the Palit over the same in Time Spy test 2. This is at 2151 @ 1.093v on both. So the FE does seem to score 1/3rd FPS higher at least in this test but it is riddled with pwr which IMO transfers onto the screen. Do you know of a good way to test frame time? Because I certainly prefer the green free Palit GPUz graphs.

You see what I have been alluding to in my previous post? You're scoring better with stock bios and it's "power limited" for just small segments in a power hungry benchmark. So what happens when you actually game with it?

Do you want to have good performance in purple circled areas where you'd be gaming? Or slightly better results in the green shaded areas that won't happen in gaming but pop up every once in awhile during benchmarks?





For gaming purposes, it's a no brainier to me, stick with stock and just game at better performance.


Edited by SlimJ87D - 5/1/17 at 2:12pm
post #230 of 333
Quote:
Originally Posted by SlimJ87D View Post

You see what I have been alluding to in my previous post? You're scoring better with stock bios and it's "power limited" for just small segments in a power hungry benchmark. So what happens when you actually game with it?

Do you want to have good performance in purple circled areas where you'd be gaming? Or slightly better results in the green shaded areas that won't happen in gaming but pop up every once in awhile during benchmarks?





For gaming purposes, it's a no brainier to me, stick with stock and just game at better performance.


I do see what you are saying but it's hard to imagine where that 30 watts is going.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: NVIDIA
Overclock.net › Forums › Graphics Cards › NVIDIA › How To Flash A Different BIOS On Your 1080 Ti.