[BIOS Mod] RX480 -> RX580 Conversions + How-to RX470/RX480 -> RX570/RX580 - Page 96 - Overclock.net - An Overclocking Community

Forum Jump: 

[BIOS Mod] RX480 -> RX580 Conversions + How-to RX470/RX480 -> RX570/RX580

Reply
 
Thread Tools
post #951 of 1052 (permalink) Old 12-27-2018, 05:28 AM
New to Overclock.net
 
mynm's Avatar
 
Join Date: Jan 2016
Posts: 623
Rep: 54 (Unique: 20)
Quote: Originally Posted by datspike View Post
I just wanted to let you all know that you can actually set those last 2 bytes (8A 00 EB FF FF FF xx xx) as you wish.
For example you want 1.1125 vcore VID under load:
1112mV (just a tad lower than 1.1125) x4 > 4448 > 1160 > 60 11
Yes, I think that is working like you say acording with the tests I did whit my broken 380, like I said here: https://www.overclock.net/forum/27689100-post841.html.

But maybe is better to use this list of svid2 voltages https://www.overclock.net/photopost/...2Voltages.jpeg

I have now a 580 nitro+, I will test it to see how is working with it.
mynm is offline  
Sponsored Links
Advertisement
 
post #952 of 1052 (permalink) Old 12-27-2018, 06:02 PM
Extreme AIR Overclocker
 
rootmoto's Avatar
 
Join Date: Feb 2014
Location: Australia
Posts: 142
Rep: 10 (Unique: 10)
@mynm I did the third party voltage controller thingy and it didn't expose a 2nd loop for MVDDC. It should be the same I2C address as the one in the Gigabyte R9 270 cause when I ran a i2cdump, it just crashed Afterburner for me.

Main PC:
Intel Core i7 8700K @4.8GHz, Gigabyte Z370 Aorus Gaming 5 (rev 1.0) motherboard, 16GB DDR4-2400 RAM Dual Channel, Gigabyte Radeon RX 470 G1 Gaming 4GB (flashed with Gigabyte RX 580 Gaming BIOS) @1360MHz core/2000MHz GDDR5 RAM, Pioneer BDR-209DBK Blu-Ray Drive , Seagate 2TB SSHD @7200RPM, Enermax Revolution 87+ 850W PSU, Cooler Master K350 Case

HTPC System:
Intel Xeon X3450 @4.0GHz, Gigabyte GA-P55-USB3 (rev 2.0) motherboard, 8GB DDR3-1600 RAM Dual Channel, Gigabyte GTX 750 Ti Windforce OC 2GB @1437.5MHz core/1602.5MHz GDDR5 RAM, ASUS,Pioneer and TSST-Corp DVD-DL Drive, Seagate 500GB HDD @7200RPM, Antec NEO ECO 450W PSU, Huntkey Case
rootmoto is offline  
post #953 of 1052 (permalink) Old 12-28-2018, 02:58 AM
New to Overclock.net
 
mynm's Avatar
 
Join Date: Jan 2016
Posts: 623
Rep: 54 (Unique: 20)
Quote: Originally Posted by rootmoto View Post
@mynm I did the third party voltage controller thingy and it didn't expose a 2nd loop for MVDDC. It should be the same I2C address as the one in the Gigabyte R9 270 cause when I ran a i2cdump, it just crashed Afterburner for me.
So voltage vddc controll is not working for you with afterburner?, or only the second loop?. I have tested afterterburner with my 580 and I can get voltage controll with the third party mode, but the second loop seems to be not working and I get system freezes using AB, so I had to uninstall it.

My 580 is using the ncp 81022 so I can't help you to test, I bought the the nitro+ because is the lower priced 580 with dual bios I saw (and driver path is not needed) and as the 380 lost was unexpected. I forget it haves the ncp 81022, the 480 nitros haves the IR one, even the lower priced 570 4gb I saw have the IR controller, but not dual bios (I don't have a igpu or a second pc and gpui), I have no idea how shappire can be using it.

Also I am having some aleatory artifacts and freezes problem after log when windows starts, and stuck core at the highest dpm after gaming. Maybe is a driver problem, I'm using the 18.12.3 driver, I have to test 18.12.1.1 to see if it solves the proble.

Also I am doing an undervolting tests using the asic profiling table, and I am getting good results . I will explain how to do it, is to long to explain.

Last edited by mynm; 12-28-2018 at 03:07 AM.
mynm is offline  
Sponsored Links
Advertisement
 
post #954 of 1052 (permalink) Old 12-28-2018, 11:17 AM
New to Overclock.net
 
datspike's Avatar
 
Join Date: Jan 2017
Posts: 206
Rep: 14 (Unique: 10)
I found out that
Code:
UCHAR  ucEnableApplyAVFS_CKS_OFF_Voltage
controls the ability to set the voltage through
Code:
USHORT usMaxVoltage_0_25mv
Must be set to zero for usMaxVoltage_0_25mv to actually do something
Spoiler!


Fun thing - my RX 590 Nitro "gaming" (1560\2100) bios which I've saved by using AtiFlash was corrupted. It was not letting the pc post from full shutdown, hanged at B2.
Turns out GPU-Z is better for bios saving, huh

R5 3600 UA1922 Auto+200
BLS2K8G4D32AESCK @ 3800C15 @ 1.46 vdram (can be lowered a bit?)
datspike is offline  
post #955 of 1052 (permalink) Old 12-28-2018, 03:15 PM - Thread Starter
hex junkie
 
hellm's Avatar
 
Join Date: Sep 2016
Location: Germany
Posts: 814
Rep: 135 (Unique: 78)
thx!

So, this is the reason why the MaxVddc and MaxVoltage_0_25mv differ in the 590 rom! The previous Polaris cards have the zero by default.

Quote: Originally Posted by mynm View Post
[..] my 380 is dead, I was trying to measure vddci and suddenly I have touched some thing accidentally with the multimeter black wire and now is dead. So I am not going to use the multimeter with a new gpu.[..]
Ouch! I guess my only GPU is also not compatible with this hisgh risk adventures.. i am sorry i gave u an idea for a way to kill your card.

..but congrats to your new Polaris card.

Comandante R.B.R.T. (Red BIOS Rebellion Team)
Polaris BIOS Mod | Vega PowerPlay Mod | MorePowerTool

Last edited by hellm; 12-28-2018 at 03:57 PM.
hellm is offline  
post #956 of 1052 (permalink) Old 12-28-2018, 03:25 PM
New to Overclock.net
 
mynm's Avatar
 
Join Date: Jan 2016
Posts: 623
Rep: 54 (Unique: 20)
Quote: Originally Posted by datspike View Post
I found out that
Code:
UCHAR  ucEnableApplyAVFS_CKS_OFF_Voltage
controls the ability to set the voltage through
Code:
USHORT usMaxVoltage_0_25mv
Must be set to zero for usMaxVoltage_0_25mv to actually do something
Spoiler!


Fun thing - my RX 590 Nitro "gaming" (1560\2100) bios which I've saved by using AtiFlash was corrupted. It was not letting the pc post from full shutdown, hanged at B2.
Turns out GPU-Z is better for bios saving, huh
Interesting, but what is this doing?. I found that AVFS is some tingh related to power effiiency: https://www.extremetech.com/gaming/2...-25x20-project . About cks, I don't know what it means, I remember to test some relates values about it into the powerplay table and if I remember correctly with it enablet I was guetting highes voltages.

I have tested also an undervolting to my 580 nitro. And I can get for now 1340 with 1.093v, I have no idea if it is totally stable or not.

My stock voltages are the ones of the first attached image. Is extrange because the voltage is 1.15v from dpm 4 to 7.

My stock Asic Profiling table is the second attached image.

I have highlighted some values I had used on it . I don't have used ulaNoCalcVddcPerDPM, but I think it is relted to ulEvvNoCalcVddc.

My modded Asic Profiling table is the third attached image.

I have changed:


ulEfuseMin to 1500: inceasing this value is decreasing the 1 to 7 dpm voltages

ulmaxVddc to 1.093v and usMaxVoltage to 1.093, without changing them even if dpm7 voltage is 1.093, the voltage is going higher up to 1.15v if wattman voltages are setted in auto. The same happens with the stock values, even if dpm7 is 1.15v the voltages can go up to 1.175v.

ulEvvDefaultVddc to 1.093 and ulEvvNoCalcVddc to 1.093, I have changed boht because if not the dpm voltage is 1.15v, but I have no idea if is needed to change both voltages.

The fourth image is the result, I had to lower the dpm 2 core mhz to 800 because if not it was no stable.

With the stock bios I am getting a 14763 firestrike graphics core, and with the mod I get 14977, because the stock bios gpu was power throttling.


Quote: Originally Posted by hellm View Post

Ouch! I guess my only GPU is also not compatible with this hisgh risk adventures.. i am sorry i gave u an idea for a way to kill your card.

..but congrats to your new Polaris card.
Thanks. Don't worry at all, is my fault, I din't have to test more things about the vddci, I was testing some mod to the Ids and the mask, but without success. I had no thought that test the gpu with the multimeter was so risky.
Attached Thumbnails
Click image for larger version

Name:	580 stock.png
Views:	19
Size:	23.4 KB
ID:	241990  

Click image for larger version

Name:	580 asic prof stock.png
Views:	48
Size:	38.6 KB
ID:	241992  

Click image for larger version

Name:	580 asic prof mod.png
Views:	35
Size:	34.7 KB
ID:	241998  

Click image for larger version

Name:	580 1.093v mod.png
Views:	24
Size:	23.6 KB
ID:	242002  


Last edited by mynm; 12-30-2018 at 04:14 PM.
mynm is offline  
post #957 of 1052 (permalink) Old 12-28-2018, 05:16 PM - Thread Starter
hex junkie
 
hellm's Avatar
 
Join Date: Sep 2016
Location: Germany
Posts: 814
Rep: 135 (Unique: 78)
thx, i will link this info in post#3.
I guess i should have linked a lot more.. will do some changes next year, and now that mynm has a Polaris we might see more info very soon anyway..
..and on the NCP81022. There is a data sheet.. https://media.digikey.com/pdf/Data%2...s/NCP81022.pdf

Comandante R.B.R.T. (Red BIOS Rebellion Team)
Polaris BIOS Mod | Vega PowerPlay Mod | MorePowerTool
hellm is offline  
post #958 of 1052 (permalink) Old 12-28-2018, 06:36 PM
New to Overclock.net
 
Merutsu's Avatar
 
Join Date: Oct 2017
Posts: 34
Rep: 1 (Unique: 1)
Quote: Originally Posted by mynm View Post
I have tested also an undervolting to my 580 nitro. And I can get for now 1340 with 1.093v, I have no idea if it is totally stable or not.
1.093v should be fine for ~1400mhz core. For 1340mhz you can lower voltage to ~1.050v
Merutsu is offline  
post #959 of 1052 (permalink) Old 12-29-2018, 02:42 AM
New to Overclock.net
 
mynm's Avatar
 
Join Date: Jan 2016
Posts: 623
Rep: 54 (Unique: 20)
Quote: Originally Posted by hellm View Post
thx, i will link this info in post#3.
I guess i should have linked a lot more.. will do some changes next year, and now that mynm has a Polaris we might see more info very soon anyway..
..and on the NCP81022. There is a data sheet.. https://media.digikey.com/pdf/Data%2...s/NCP81022.pdf
You are welcome.

I'm having some problems with my gpu one is that after gaming some times clock is stuck at the max dmp like here: https://community.amd.com/thread/222777 , that is a one year old report so seems to not be solved...

The thing is that to solve it I have to restart the pc and after login I get artifacts and a VIDEO_TDR_FAILURE Atikmpag.sys blue screen like here: https://community.amd.com/thread/230386

But I am not sure if the two problesm are related.

I have tested 18.12.3 and 18.12.1.1, but only using DDU to unistall.

Now I am testing this: "Then uninstall all AMD driver software using Windows Uninstall. Now download and run DDU (Display Driver Uninstaller) in safe mode. This will completely eliminate any leftover AMD driver software or files. Delete if you have C:\AMD folder in the Root Directory. reboot." like is say in the link.

I am using the stock bios to be sure that is not my modded bios problem. But I remember to had the problem with the stock bios after intalling afterburner so I was thinking afterburner are the guilty. And then had installes trixx with no problem, but maybe it is using some thing into windows register, because the gpu have some led that can be configured, so maybe it can cause some problem as well.

If the problem is not solver I will have to return the gpu to the store. But maybe I have to do a clean windows instalation before to test if a problem with older driver.

Quote: Originally Posted by Merutsu View Post
1.093v should be fine for ~1400mhz core. For 1340mhz you can lower voltage to ~1.050v
I had tested 1380 with 1.093v and was no stable, so I'm not lucky .

The thing is that with my 380 I saw a point where bios core clock can't be increase without needing 50mv more than with software. The point was between 1040 and 1075 mhz I don't remember were exactly and with 1040mhz the stock bios was using the max asci profiling voltage 1.225, and going for more than that I was seeing 1.225 for lower dpm, like I see here with the 580 of using 1.15v for lower dpms. I have to test if lowering the clocks I am getting lower voltages than 1.15v for the lower dpms and if I can do better ocing with software.
mynm is offline  
post #960 of 1052 (permalink) Old 12-29-2018, 05:57 PM
New to Overclock.net
 
Merutsu's Avatar
 
Join Date: Oct 2017
Posts: 34
Rep: 1 (Unique: 1)
Quote: Originally Posted by mynm View Post
The fourth image is the result, I had to lower the dpm 2 core mhz to 800 because if not it was no stable.
I remember that my nitro rx580 has -25mV offset for DPM1-6 states with stock bios. Did you remove these offsets?

Quote: Originally Posted by mynm View Post
I had tested 1380 with 1.093v and was no stable, so I'm not lucky .
what vddc gpu-z shows under 100% gpu load?
Merutsu is offline  
Reply

Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the Overclock.net - An Overclocking Community forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
If you do not want to register, fill this field only and the name will be used as user name for your post.
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:

Log-in



Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off