AGESA FW stack patched bioses for 3rd gen - Page 54 - Overclock.net - An Overclocking Community

Forum Jump: 

AGESA FW stack patched bioses for 3rd gen

Reply
 
Thread Tools
post #531 of 539 (permalink) Old 11-22-2019, 10:37 PM
New to Overclock.net
 
Synoxia's Avatar
 
Join Date: Apr 2017
Posts: 1,011
Rep: 7 (Unique: 5)
Honestly after seeing performance degradation on 1.0.0.4b i think i will just skip to 1.0.0.5 :/

Tryndamere
(13 items)
CPU
AMD Ryzen 3700x
Motherboard
Asus C7H Hero Wifi
GPU
Radeon Vega 64
RAM
G-Skill Tridentz OC 16 gb 3800 c14-16-17-16-32-48
Power Supply
EVGA G3 Gold
Cooling
Noctua D15
Case
Phanteks Enthoo pro TG
Operating System
Win 10 1903 build 18950 64bit
Monitor
Viewsonic XG2401
Keyboard
Fnatic RGB Cherry mx brown
Mouse
Logitech G Pro Wireless 950 dpi
Mousepad
Artisan ZERO Fx
Audio
JBL305
▲ hide details ▲
Synoxia is offline  
Sponsored Links
Advertisement
 
post #532 of 539 (permalink) Old 11-22-2019, 10:46 PM
New to Overclock.net
 
VPII's Avatar
 
Join Date: Dec 2014
Location: Hout Bay, Cape Town, South Africa
Posts: 711
Rep: 9 (Unique: 8)
Quote: Originally Posted by Synoxia View Post
Honestly after seeing performance degradation on 1.0.0.4b i think i will just skip to 1.0.0.5 :/
Thank you, I felt like I was the only person seeing this and got some people going off at me because I was comparing small differences, but hell the differences are there to begin with.
VPII is offline  
post #533 of 539 (permalink) Old 11-25-2019, 02:28 AM
Stock is too casual~
 
Veii's Avatar
 
Join Date: Aug 2018
Location: Austria / Bulgaria
Posts: 302
Rep: 14 (Unique: 12)
Out of topic Question @The Stilt
Could you take a look into the X570 Taichi current 1004B release and maybe show some insight "how even" did ASRock made a dual bios which is P1.10 for anything 1st & 2nd gen put in - but switches to original Bios if a 3rd gen CPU is detected

Of course if you're bored - i rly wonder where they split the bios in half (location)
And wonder why there is support for Picasso 3400G , while there is no support for Raven Ridge 2400G
In the meantime 1700X works flawlessly and we have a lot of options unlocked now with it , in this combined/seperated bios mixture

Personal Records
SiSandra Inter-Core Bandwith 14nm Ryzen #2 Lost by a 101Mhz FSB OC #1 :>
RAM > Viper Steel @ 3466C14 (64.2ns) & Alt | Hynix-MFR 3333CL14 (72.5ns) & Alt

GPU > Gigabyte 970 G1 @ 1595|1900 (Hynix) | RX 580 Red Devil GS @ 1620|2200 (Micron)
CPU > TR 1950X [email protected]1.2v & [email protected]1.365v[105Mhz FSB](3.23Beta)
#Main: 1700X [email protected]1.365v1.73Ghz IMC CL14 (64.2ns)」+ Dark Rock 3 | X370 Taichi 1.0.0.3 ABBA by The Stilt | Viper Steel [PVS416G400C9K]| Gigabyte 970 G1 | Samsung 850 Evo
Veii is offline  
Sponsored Links
Advertisement
 
post #534 of 539 (permalink) Old 11-29-2019, 08:24 AM
New to Overclock.net
 
Join Date: Nov 2019
Posts: 1
Rep: 0
Quote: Originally Posted by VPII View Post
Thank you, I felt like I was the only person seeing this and got some people going off at me because I was comparing small differences, but hell the differences are there to begin with.
Dude its horrible on my x570 TUF and i cant go back/ at least mod the Bclk or disable speed spectrum. I tried flashrom but got stuck a the difference in size.

Last edited by yojalata; 11-29-2019 at 08:33 AM.
yojalata is offline  
post #535 of 539 (permalink) Old 11-29-2019, 10:27 AM - Thread Starter
New to Overclock.net
 
The Stilt's Avatar
 
Join Date: May 2013
Posts: 2,574
Rep: 861 (Unique: 329)
Quote: Originally Posted by Veii View Post
Out of topic Question @The Stilt
Could you take a look into the X570 Taichi current 1004B release and maybe show some insight "how even" did ASRock made a dual bios which is P1.10 for anything 1st & 2nd gen put in - but switches to original Bios if a 3rd gen CPU is detected

Of course if you're bored - i rly wonder where they split the bios in half (location)
And wonder why there is support for Picasso 3400G , while there is no support for Raven Ridge 2400G
In the meantime 1700X works flawlessly and we have a lot of options unlocked now with it , in this combined/seperated bios mixture
This is a standard functionality.
Basically the bios acts like it is two different bioses located on the same chip.
The bios image contains an addressing table, which is read by PSP (IIRC). Then depending on the installed CPU, it loads either the lower or upper part of the flash.
On most motherboard the first half is the 1st and 2nd gen. stuff, whereas the upper part is for the 3rd gen.
The Stilt is offline  
post #536 of 539 (permalink) Old 11-29-2019, 04:00 PM
New to Overclock.net
 
Hammerfest's Avatar
 
Join Date: Jul 2012
Posts: 18
Rep: 0
Quote: Originally Posted by The Stilt View Post
This is a standard functionality.
Basically the bios acts like it is two different bioses located on the same chip.
The bios image contains an addressing table, which is read by PSP (IIRC). Then depending on the installed CPU, it loads either the lower or upper part of the flash.
On most motherboard the first half is the 1st and 2nd gen. stuff, whereas the upper part is for the 3rd gen.

True, I get his confusion regarding the 3400G however, seeing as it loads Gen2 "BIOS", even though its not Gen2, they are treating it as such and probably just applying optimizations as needed since its 3000 "series".

Which is fine by me at least.


Thanksgiving week has turned out to be a great Holiday!

Of my boards/systems:
A300W: Still out in the cold (commmmonnnn, I have 3 of these and want more, just needs the love!)
B450M Steel Legend: Got the love on Nov 28th
B450 Pro4: Got BETA love on Nov 27th
ROG STRIX B450-F Gaming: Got BETA love on Nov 20th
GA-AB350N-Gaming WiFi: Got the love on Nov 27th
X470 Aorus Gaming 7 WIFI: Got the love Nov 27th

I have a good 5 other boards in system's I directly interact with, and a stupid amount that I do support for, but so far, good start to the holidays, the boot speed is greatly appreciated!


Question, how can I check if they have updated all modules such as SMU and the rest, I can already tell they didnt update the non-AMD modules from UBU, but am at a loss to see the rest.
Maybe you can work with UBU creator to automate SMU and other AMD modules like he did for Intel eons ago since it seems AMD is much easier to streamline now then PRE Rizen.
Hammerfest is offline  
post #537 of 539 (permalink) Old 12-01-2019, 06:36 PM
New to Overclock.net
 
mat9v's Avatar
 
Join Date: May 2014
Posts: 426
Rep: 10 (Unique: 8)
Quick question - if I have already flashed a bios based on AGESA 1.0.0.4, can I safely flash it back to 1.0.0.3 ABBA?
If yes, could I request an Asus X570 Prime Pro bios?
https://dlcdnets.asus.com/pub/ASUS/m...-ASUS-1201.zip
1.0.0.4B gives me significantly lower max boost clocks. While it gives compatibility with CPPC, performance results are lower by few % (4625Mhz vs 4517Mhz) and multicore CB20 test drops (with PBO enabled) from 7475 to 73xx points.


Last edited by mat9v; 12-01-2019 at 06:43 PM.
mat9v is offline  
post #538 of 539 (permalink) Old 12-02-2019, 06:17 AM
Stock is too casual~
 
Veii's Avatar
 
Join Date: Aug 2018
Location: Austria / Bulgaria
Posts: 302
Rep: 14 (Unique: 12)
Quote: Originally Posted by mat9v View Post
Quick question - if I have already flashed a bios based on AGESA 1.0.0.4, can I safely flash it back to 1.0.0.3 ABBA?
If yes, could I request an Asus X570 Prime Pro bios?
https://dlcdnets.asus.com/pub/ASUS/m...-ASUS-1201.zip
1.0.0.4B gives me significantly lower max boost clocks. While it gives compatibility with CPPC, performance results are lower by few % (4625Mhz vs 4517Mhz) and multicore CB20 test drops (with PBO enabled) from 7475 to 73xx points.
You can flash back without any problems
the Flashtool fully wipes every trace and writes the new bios
There are situations where boards with USB flashback support - should focus on USB flashing
But the way this tools works is - a full clean SPI flash / nearly identical to external SPI flashers
Soo if you want, you could even downgrade to the first factory bios in existence

About ABBA vs 1.0.0.4
As The Stilt mentioned back in Post #248
Spoiler!

Performance issues with 1004B are visual , SMU did change BUT - as we know from before the difference between updated ABBA and 1004 is identical when it comes to any kind of boosting behaviour
What does change, are many fixes on the later 1004B and quite a few new bugs too
EDC is bugged
SOC voltage on some boards is fixed
PBO (again EDC) values are broken and don't accept HEX or mixed of Decimal and Hex

Overall, the biggest difference you see is prefered cores - which can be bugged and HWInfo (non beta) doesn't know what to read out
There should be no boosting difference in workload performance scale - but there can be very likely boosting difference in the numeral state, just wrong readout from the wrong prefered cores / in comparison to before unknown CPPC , making boosting behaviour be less accurate and less effective

That aside,
You can not inject 1004B fixes into ABBA bioses - different modules, different license requirements, just won't work
You could downgrade as you mentioned to ABBA, still enjoying the current SMU - but you will lose accuracy and only visually get a wrong boosting number representation
Not to forget, you can not use 1usmus's powerplan without CPPC and upcomming windows sheduling fixes will be worthless for you without CPPC
Your performance difference may be bugged PBO, bugged applied voltage for CLD0 and SOC and also just wrong readout or too old windows version without propper utilisation of CPPC

Personal Records
SiSandra Inter-Core Bandwith 14nm Ryzen #2 Lost by a 101Mhz FSB OC #1 :>
RAM > Viper Steel @ 3466C14 (64.2ns) & Alt | Hynix-MFR 3333CL14 (72.5ns) & Alt

GPU > Gigabyte 970 G1 @ 1595|1900 (Hynix) | RX 580 Red Devil GS @ 1620|2200 (Micron)
CPU > TR 1950X [email protected]1.2v & [email protected]1.365v[105Mhz FSB](3.23Beta)
#Main: 1700X [email protected]1.365v1.73Ghz IMC CL14 (64.2ns)」+ Dark Rock 3 | X370 Taichi 1.0.0.3 ABBA by The Stilt | Viper Steel [PVS416G400C9K]| Gigabyte 970 G1 | Samsung 850 Evo
Veii is offline  
post #539 of 539 (permalink) Old 12-02-2019, 05:12 PM
New to Overclock.net
 
mat9v's Avatar
 
Join Date: May 2014
Posts: 426
Rep: 10 (Unique: 8)
Quote: Originally Posted by Veii View Post
You can flash back without any problems
the Flashtool fully wipes every trace and writes the new bios
There are situations where boards with USB flashback support - should focus on USB flashing
But the way this tools works is - a full clean SPI flash / nearly identical to external SPI flashers
Soo if you want, you could even downgrade to the first factory bios in existence

About ABBA vs 1.0.0.4
As The Stilt mentioned back in Post #248

Performance issues with 1004B are visual , SMU did change BUT - as we know from before the difference between updated ABBA and 1004 is identical when it comes to any kind of boosting behaviour
What does change, are many fixes on the later 1004B and quite a few new bugs too
EDC is bugged
SOC voltage on some boards is fixed
PBO (again EDC) values are broken and don't accept HEX or mixed of Decimal and Hex

Overall, the biggest difference you see is prefered cores - which can be bugged and HWInfo (non beta) doesn't know what to read out
There should be no boosting difference in workload performance scale - but there can be very likely boosting difference in the numeral state, just wrong readout from the wrong prefered cores / in comparison to before unknown CPPC , making boosting behaviour be less accurate and less effective

That aside,
You can not inject 1004B fixes into ABBA bioses - different modules, different license requirements, just won't work
You could downgrade as you mentioned to ABBA, still enjoying the current SMU - but you will lose accuracy and only visually get a wrong boosting number representation
Not to forget, you can not use 1usmus's powerplan without CPPC and upcomming windows sheduling fixes will be worthless for you without CPPC
Your performance difference may be bugged PBO, bugged applied voltage for CLD0 and SOC and also just wrong readout or too old windows version without propper utilisation of CPPC
In order
1. Good to know I can flash it back safely. Can you point me to some tool that can convert the bios file that is posted on Asus page to one that Flashtool can actually safely flash? I remember reading something about converting .cap files to .bin but I don't want to make a mistake.
2. Now to 1.0.0.4B in my case - I have of course the latest beta HWiNFO - the reported clocks are lower, the PERFORMANCE in all apps is lower, PBO is higher then stock but lower then on 1.0.0.3 ABBA. All tests done in clean system with only drivers and testing apps present and Antivirus/Indexing disabled. Please believe me, I would not cry wolf if I didn't test it extensively (I mean stock, PBO and autoOC) and while I make no claims about other bioses or motherboards in this one case it sux.
3. As to the preferred cores, I don't really care as I use ProcessLasso to force apps to use selected cores and it does not matter if the game or an app run on CCX0 or CCX1 on 3900X since they are of similar quality - the 75-100Mhz higher boost is more important. In fact all cores in CCX0 and 1 are able to reach the same frequency with seemingly same 1.47/1.5V voltage. Oh yes, one other thing, all tests are done with all fans at full speed and temps below 60C in case of ST tests and below 70C in case of MT tests. Even at stock, without PBO on completely clean system with fans at maximum and 52C temp under load for ST scenario I can't reach even 4550Mhz (without any active app, without ProcessLasso to muddle the waters). Others in such a case report up to 4650Mhz on ABBA bios and before 1.0.0.4 I was routinely getting 4625Mhz with performance numbers agreeing.
4. "You could downgrade as you mentioned to ABBA, still enjoying the current SMU" - you mean that if I flash 1.0.0.3 ABBA bios I will still have the new SMU? That I will get falsely high CPU clocks and still get low performance? That is the scenario I do not want to have
5. Oh, and I am running W10 1909, latest drivers and so on, been doing that (beta) before even flashing 1.0.0.4 bios.

mat9v is offline  
Reply

Tags
am4 , bios flash , ryzen

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