• Save 15% on ALL SFF Network merch, until Dec 31st! Use code SFF2024 at checkout. Click here!

Motherboard Group Buy/Crowdfunded Modded BIOS for the Asrock A300m (Deskmini A300)

Hammerfest

Average Stuffer
Jul 15, 2019
55
43
Because that's a known buggy beta BIOS, and lacks the clock and voltage adjustments of the X300?

Uhhh wut? Known buggy it is not.
There is one person who said they had an issue with XMP profile, but XMP profiles are hardly cross platform friendly.
No issues for myself or countless others in the forum who have flashed it so far.
 

W4RR10R

Cable-Tie Ninja
Original poster
Jan 29, 2019
211
211
So I still haven't tried it, but "X3MSTX_1.40" UEFI isn't recognized by the instant flash update tool, it doesn't even show up, but if you change the X to an A, A3MSTX1.40 it shows up as an option, but I haven't clicked the "update" button.
 

fadsarmy

Caliper Novice
Oct 24, 2017
32
12
I wanted to throw out some positive virtualization news with the new BIOSes.
I currently have 3.60H on my A300W with 3400G and the USB passthrough to KVM windows 7 guest works perfectly just like on my other Asrock AM4 machines.
The back USB and side USB ports are 1 controller, the 2 front USB ports are a separate controller that I tested passthrough.

Also I haven't had any trouble yet detecting/booting the ssd with my Silicon Power 512GB NVMe drive. (note that on my other machines with 1TB and 2TB versions of this model, they have no trouble handling 101BLCK, even when I RAID0 4 of them together with PLX card on X370 :-))

However I have had 2 issues with the unit so far.
When I boot with at least certain flash drives plugged into the front ports (example exfat formated usb3 drive), when I enter the BIOS the mouse and keyboard don't work. I have never encountered this issue on other Asrock mobos.
It boots so fast it is hard to get into the BIOS menu half of the time.
Disable fast boot.
 

johnyb0y

Efficiency Noob
Aug 3, 2020
5
0
It boots so fast it is hard to get into the BIOS menu half of the time.

AsRock has a "Restart to UEFI Tool" which you can use to get into the Bios as well. On the A300 page under Downloads:
 

gustav

Cable-Tie Ninja
Jun 18, 2020
193
90
So I still haven't tried it, but "X3MSTX_1.40" UEFI isn't recognized by the instant flash update tool, it doesn't even show up, but if you change the X to an A, A3MSTX1.40 it shows up as an option, but I haven't clicked the "update" button.
Very odd, their Flash tool decides for compatability, wheather the file starts with an A... Untrustworthy
 

catawalks

Average Stuffer
Dec 17, 2019
62
88
Very odd, their Flash tool decides for compatability, wheather the file starts with an A... Untrustworthy
It's probably not whether is starts with just an "A" or not, but rather it's looking for a specific filename format for this specific motherboard model. It most likely won't recognize any file unless it starts with "A3MSTX" and then it will do more checks from there.
 

gustav

Cable-Tie Ninja
Jun 18, 2020
193
90
Yea, I mean it's BIOS they could have used a header in the file itself. Or somehow do a CRC computation, which value is also provided in the .hex to compare
 

catawalks

Average Stuffer
Dec 17, 2019
62
88
Yea, I mean it's BIOS they could have used a header in the file itself. Or somehow do a CRC computation, which value is also provided in the .hex to compare

Those checks are most likely there. The above poster didn't actually try to apply the BIOS. Most likely when the BIOS is actually applied those checks will run. But the first step before the other checks are run is to make sure the filename starts with the appropriate name for the board.
 

takisha_wan

Efficiency Noob
Jul 28, 2020
7
3
Those checks are most likely there. The above poster didn't actually try to apply the BIOS. Most likely when the BIOS is actually applied those checks will run. But the first step before the other checks are run is to make sure the filename starts with the appropriate name for the board.
There is a report from Hong Kong stating that just change the file name is not sufficient. The Instant Flash tool would give a invalid response.
 
Last edited:

yuusou

SFF Lingo Aficionado
Mar 16, 2019
118
75
Would be interesting if someone with a flasher could try doing so directly, if that would effectively change the board to an X300.
 

takisha_wan

Efficiency Noob
Jul 28, 2020
7
3
Would be interesting if someone with a flasher could try doing so directly, if that would effectively change the board to an X300.
I do have AFUWIN windows flash application. But now I'm using A300 as my work machine and I don't have other spare PC around, so I can't take the risk of brick my work machine.
If there are any brave (and reckless maybe?) souls, here is the AFUWIN application with X300 BIOS file. Caution! I haven't tried to flash this before, so do it on your own risk. I'm not responsible for damage of loss caused by any user try to do this and brick their machine.
 
Last edited:

duha21

Chassis Packer
May 27, 2020
18
19

TomHC81

Trash Compacter
May 21, 2020
44
19
www.sitepoint.sk
I bought tray 3400G for 75€ and can't find a good reason to spend ~200€ for 6C/12T which I basically don't need badly (except the fact, it would be nice to have). Maybe a x-mas gift? ?
 

gustav

Cable-Tie Ninja
Jun 18, 2020
193
90
So, today I've tested the 3.60N with 2400G. Successful boot. But stability issues running RAM@XMP (3200). (Against the SoC voltage, which you are not anymore able to change)


Rolled back to 3.60K. Same behavior with RAM@XMP but I'm able to adjust SoC VID. Stability recovered. Staying @3.60K.
3.60K still reporting errors in IOMMU and dmesg on Ubuntu 20.04.

Hoping to be able to adjust SoC VID in official AGESA v2.
Update on 3.60R: Working very good on 2400G and 3200MHz RAM. Due to added SoC VID selection (in version before) I can now run RAM@XMP without stability issues (SoC VID=1.11875 (VID=1.12500 even better/less artifacts)). Vega has additionaly ~3-4fps more. (See Screenshot) All-in-all very good support, very good revision! In terms of flexibility. Gonna stay on 3.60R

With VID=1.125V under full load (FurMark + 5 threads CPU burn-in) the SoC VID breaks down to VID=1.031V - which functions well. No stability issues right now. Renoir can come. Thanks @ASRock System !!

@TomHC81 @johnyb0y I wanted to add, that the CPU performace of the 4650G(~210€, German Store) is about 40-50% higher compared to Ryzen 2400G (see Cinebench R15, R20). It seems to be worth of an upgrade. ( My opinion :) )


If you want to run it yourself:

Render:Direct3D11
Mode:1920x1080 fullscreen
PresetCustom: API DX11, Qualy: Low, Stereo 3D off, AA off
QualityLow
 
Last edited:
  • Like
Reactions: Curiosity

bowylechat

Efficiency Noob
Aug 12, 2020
6
4
Update on 3.60R: Working very good on 2400G and 3200MHz RAM. Due to added SoC VID selection (in version before) I can now run RAM@XMP without stability issues (SoC VID=1.11875 (VID=1.12500 even better/less artifacts)). Vega has additionaly ~3-4fps more. (See Screenshot) All-in-all very good support, very good revision! In terms of flexibility. Gonna stay on 3.60R

With VID=1.125V under full load (FurMark + 5 threads CPU burn-in) the SoC VID breaks down to VID=1.031V - which functions well. No stability issues right now. Renoir can come. Thanks @ASRock System !!

@TomHC81 @johnyb0y I wanted to add, that the CPU performace of the 4650G(~210€, German Store) is about 40-50% higher compared to Ryzen 2400G (see Cinebench R15, R20). It seems to be woth of an upgrade. ( My opinion :) )


If you want to run it yourself:

Render:Direct3D11
Mode:1920x1080 fullscreen
PresetCustom: API DX11, Qualy: Low, Stereo 3D off, AA off
QualityLow

Thanks for your return !
I've tested 3.60R with ddr4 3200 pushed to 3266 with ryzen 5 3400G and VID:1.125.
Everythings were fine.
Waiting for Renoir too.
 
  • Like
Reactions: gustav

gustav

Cable-Tie Ninja
Jun 18, 2020
193
90
Thanks for your return !
I've tested 3.60R with ddr4 3200 pushed to 3266 with ryzen 5 3400G and VID:1.125.
Everythings were fine.
Waiting for Renoir too.
Thanks for your kind words, I had to bump up my VID to 1.13750V. My system hung during a game.
Afterwards I remember - and I suggest to anyone - to benchmark the system using FurMark CPU Burner with 5 Threads, start it, and afterwards UNIGINE Valley.
System does not hang anymore. 39-40 fps with 5 threads @ 100%.

Othe method is the built-in GTA V benchmark. It's about conccurency on the Memory I/O due to access by the CPU and APU simultaniously.
The load is very specific, but it is indeed overall stability check.

This may also be the improvement in terms of the Memory I/O of Ryzen 2000G vs 3000G Series. Yours is fine with less VID, since it's like Revision 2. Very nice. I wonder about Renoir. :)

Update #1: There was one time my system did BSOD on VID=1.13750V. I added one increment in the voltage, but since this is way too much, I returned instantly back to VID=1.13750V. No crashes so far. Just wanted to let you know, maybe my 2400G is not the best sample :)

Also, if I'd like to got with the new Gen. Renoirs, then the 4650G would be the right fit in terms of CPU power. But we are on A300, and only the 4750G can provide the full Vega-Experience. Damn... I think I will go that way. As of today, you can buy the 4650G without any troubles here in EU/GER.
 
Last edited:
  • Like
Reactions: Curiosity

blackreaver

What's an ITX?
New User
Aug 21, 2020
1
0
Does anyone know what the exact changes were from BIOS 3.50 to 3.60 on the A300?
ASRock site only says "Improve system compatibility" .

Having a couple of issues with occasional random crashes and peripherals not being recognised on boot and wondering whether a BIOS upgrade will solve.

Running:

Linux Mint 20 Ulyana | DeskMini A300 with A8-9600