Quick update: I managed to swap in the Zen 2 microcodes and flash the new BIOS to the motherboard. It still boots up on the current APU so I didn't break anything.
I think the next step before plugging in a Zen 2 processor is to get my eGPU up and running on the motherboard. My first attempt with the ADT.Link R43SL cable didn't power up the eGPU. I'm going to troubleshoot that on my main PC's m.2 slot and see if I get better results.
I'm willing to share the BIOS code with anyone else who want's to trial it & may already have an eGPU running, but you'll need a CH341A Programmer with a 1.8V adapter (~$10 USD) to flash the BIOS chip... there's no other way to flash a modded BIOS to an ASRock motherboard right now; their security protection prevents "Instant Flash". The process is very easy though.
I'm also reading up on how to replace the AGESA code with a better version. Stay tuned!
I think the next step before plugging in a Zen 2 processor is to get my eGPU up and running on the motherboard. My first attempt with the ADT.Link R43SL cable didn't power up the eGPU. I'm going to troubleshoot that on my main PC's m.2 slot and see if I get better results.
I'm willing to share the BIOS code with anyone else who want's to trial it & may already have an eGPU running, but you'll need a CH341A Programmer with a 1.8V adapter (~$10 USD) to flash the BIOS chip... there's no other way to flash a modded BIOS to an ASRock motherboard right now; their security protection prevents "Instant Flash". The process is very easy though.
I'm also reading up on how to replace the AGESA code with a better version. Stay tuned!
Last edited:
There are two Combo-AM4 1.0.0.3 based beta BIOSes out there, L3.51 and L3.53. The former you can grab at https://shop.jzelectronic.de/news.php?id=1576587840&sw= (dropdown box at bottom of page "BIOS Downloads Auswahl"), the latter from the link in http://forum.asrock.com/forum_posts.asp?TID=12597&title=solveddeskmini-a300-hyperv-boot-failure
The SMU code is still old in these but the PSP block looks less messy than in the production BIOSes so may give an easier start.
@ConsolidatedResults Is this a german modding forum? People are just creating beta BIOSes for ASRock systems? It's cool & I wish I spoke German because they seem to know a lot more than I do, although none of the A300 BIOSes appear to have AGESA 1.0.0.4...maybe intentionally to limit support. My plan was to steal AGESA 1.0.0.4 from the x570 Taichi BIOS assuming I can figure it out. I'm also surprised about the disclaimer on Bristol Ridge CPUs being unsupported & limited BIOS size. I wonder what's limiting space on the beta BIOSes.
In other new, I finally got my eGPU (5700XT) working; will need to boot up a few more times to see if the bugs are gone. One small step at a time.
In other new, I finally got my eGPU (5700XT) working; will need to boot up a few more times to see if the bugs are gone. One small step at a time.
Anandtech did a write up of this issue here. Essentially, the AM4 socket has been around so long and supported so many CPUs, that the BIOS chips are too small to hold microcode for all of them. So to make everything fit, when they add new generations of CPUs to a BIOS they sometimes have to drop the oldest supported generation due to size limits.
It's a retailer who gets limited release / beta BIOS versions from ASRock directly. Sometimes he will host versions that ASRock does not (yet) link / host on their website.
Basically stumped right now. I took the L3.51 BIOS (AGESA 1.0.0.3) from @ConsolidatedResults link above and dropped in the Zen 2 microcodes but the system would not boot with a Ryzen 3600 (...boots fine with the A8-9600). After pressing power, the CPU fan spins up but the motherboard doesn't seem to distribute power to anything else (usb ports, m.2, etc). There is a 4-pin speaker header on the board, I'll need a speaker but maybe there are post codes I can listen for?? I'm assuming it's some kind of internal BIOS check preventing power on, maybe that can be disabled. AMD BIOSes & AGESA are a like a black box to the modding community; not much written about them.
As with my last post, I am not able to edit the BIOS to support Zen 2 processors. I have looked across several forums, and asked questions, and there is just very little knowledge of AMD AMI BIOSes. One can easily look at the structure of the BIOS (lots of modules) but there is no knowledge of how the parts interact. My poor motherboard has been brain-wiped and flatlined more times than I honestly know.
Short of brute force testing of code changes, a solution can only be provided by the motherboard manufacturer. I have reached out to ASRock a half dozen times asking for a BIOS and the only response I've gotten is that Zen 2 cannot work because there is no PCIe slot. That's not completely true as I can run a GPU via the m.2 port and a riser. I can even install drivers for the GPU.
I encourage anyone who would like the A300M-STX motherboard to support Zen 2 processors to contact ASRock technical support (asrock_tsd@asrock.com.tw) and make this request yourself. By chance, maybe enough voices will get the result we want.
Short of brute force testing of code changes, a solution can only be provided by the motherboard manufacturer. I have reached out to ASRock a half dozen times asking for a BIOS and the only response I've gotten is that Zen 2 cannot work because there is no PCIe slot. That's not completely true as I can run a GPU via the m.2 port and a riser. I can even install drivers for the GPU.
I encourage anyone who would like the A300M-STX motherboard to support Zen 2 processors to contact ASRock technical support (asrock_tsd@asrock.com.tw) and make this request yourself. By chance, maybe enough voices will get the result we want.
Thank you for working this issue up to now. I'm sorry to hear that you've reached an impasse. Regarding ASRock, other forum members have contacted the manufacturer on several occasions with no results. Unfortunately, I suspect that this version of the A300 is moving towards unsupported status and ASRock is unlikely to spare additional resources on it's development. Chances are that any progress towards POSTing with a non-APU Ryzen CPU will come from community devs.
As for whether it'll reach unsupported status, we'll know more when the new Renoir chips launch (hopefully will be announced at CES this week) and whether the A300stx board gets a BIOS update to support those.
I posted in the other main A300W thread, but I'll mention it here as well... I'm really exited at the prospect of an A300W with a 4800G 8C/16T APU. Hopefully the APU will be released by AMD and ASRock will offer support. I guess time will tell, waiting sux sometimes.
"
AGESA Combo-AM4 1.0.0.1 BIOS can fully supports Picasso/Raven Ridge/Bristol Ridge CPUs.
Combo-AM4 1.0.0.4 patch B BIOS for A300M-STX will not be released for now.
Thank you!
ASRock America Support
"
I can only repeat what I have said before, as long as ANYONE and EVERYONE continues to send in tickets (dont spam obviously) requesting the update (and giving the reasons, IE faster boot speed, better RAID performance which DO impact Zen/Zen+, not to mention the countless other bug/security fix's that have been made), they will then have a reason to do so, until then, they are operating in a standard "do you update your parts that sell like hot cakes, or the ones that sell here and there" business state and will continue to do so.
I can understand BIOS size being an issue, but its quite easy to make 2 BIOS release's, one for Zen/Zen+/Zen2 and one for non Zen.
Sure there will be people who dont understand and flash the wrong one, but as it stands I work for a company with equipment where people flash unsupported open source stuff all day and want to complain when they flash the wrong one, but thats what warranty policy/agreement's are for.
At the rate they are going @GLSRacer with what the reply above I got from them, this will NOT happen as it will require that they update the AGESA which they seem reluctant to do.
AGESA Combo-AM4 1.0.0.1 BIOS can fully supports Picasso/Raven Ridge/Bristol Ridge CPUs.
Combo-AM4 1.0.0.4 patch B BIOS for A300M-STX will not be released for now.
Thank you!
ASRock America Support
"
I can only repeat what I have said before, as long as ANYONE and EVERYONE continues to send in tickets (dont spam obviously) requesting the update (and giving the reasons, IE faster boot speed, better RAID performance which DO impact Zen/Zen+, not to mention the countless other bug/security fix's that have been made), they will then have a reason to do so, until then, they are operating in a standard "do you update your parts that sell like hot cakes, or the ones that sell here and there" business state and will continue to do so.
I can understand BIOS size being an issue, but its quite easy to make 2 BIOS release's, one for Zen/Zen+/Zen2 and one for non Zen.
Sure there will be people who dont understand and flash the wrong one, but as it stands I work for a company with equipment where people flash unsupported open source stuff all day and want to complain when they flash the wrong one, but thats what warranty policy/agreement's are for.
At the rate they are going @GLSRacer with what the reply above I got from them, this will NOT happen as it will require that they update the AGESA which they seem reluctant to do.
For anyone who has not submitted a ticket yet, it looks like there are actual ISSUES with Ryzen 3000/3200/3400G and AGESA 1.0.0.1
forum.asrock.com
For those of you with a backup and are on 3.60, even on Ryzen 200/2200/2400G, maybe give a test to enabling Hyper-V with the Virtualization feature on in BIOS (it supposibly will render system unbootable until BIOS Virtualization option turned off), see if the issue occurs, if it does, thats more of a valid reason to upgrade AGESA then oh look performance improvement + fix's... (and yes the irony is not lost on me that fix's are reasons but unless you have exact details no-one cares :/)
from what is reported, BETA 3.51 had 1.0.0.3 (ABB or ABBA cant tell), dunno what BETA 3.53 AGESA was, either way, issue post says problem exists in 3.60 so whatever fix happened FROM ASROCK in 3.53 didnt make it in just like no AGESA update
(Solved)Deskmini A300 Hyper-V_boot failure
tested with a few A300M with 3400G: After installing HyperV Feature, the reboot of windows keeps failing.When...
For those of you with a backup and are on 3.60, even on Ryzen 200/2200/2400G, maybe give a test to enabling Hyper-V with the Virtualization feature on in BIOS (it supposibly will render system unbootable until BIOS Virtualization option turned off), see if the issue occurs, if it does, thats more of a valid reason to upgrade AGESA then oh look performance improvement + fix's... (and yes the irony is not lost on me that fix's are reasons but unless you have exact details no-one cares :/)
from what is reported, BETA 3.51 had 1.0.0.3 (ABB or ABBA cant tell), dunno what BETA 3.53 AGESA was, either way, issue post says problem exists in 3.60 so whatever fix happened FROM ASROCK in 3.53 didnt make it in just like no AGESA update
I'm using the 2nd or 3rd BIOS with a 2400G and I have had no problems that would make me want to update. Hyper-V works just fine and I am running 2 VMs 24/7 right now. I just want to make sure we get updated to support the 4000 series desktop APUs. I'll try to send a support email to ASRock mentioning the above issue when I have a chance
I asked:
"Hello!
Can you provide a BIOS update to AMD AGESA Combo-AM4 1.0.0.4 for the A300 DeskMini?
If not, do you have an estimate of when it will be available?
Thank you!"
This is the answer i got:
"Hello,
have to say sorry, I do not have an ETA when this bios is available.
I know. The bios department have it on Schedule as soon as possible.
Best regards,
ASRock Support
ASRock Europe B.V.
Bijsterhuizen 1111
6546 AR Nijmegen
The Netherlands
www.asrock.com"
In that case, I like the Netherlands better then the US... even if that reply reaks of BS... but maybe thats just my Murican thought pattern speaking... well see! I pray to be shown incorrect.
