Form factor doesn't matter to me, its the only x300 board that I've seen so far that supports non-APUs. Plus, since it's ASRock, swapping UEFI modules should not be as hard.
-
Save 15% on ALL SFF Network merch, until Dec 31st! Use code SFF2024 at checkout. Click here!
If form factor doesn't matter I'm curious: which advantage does a X300 motherboard have compared to e.g. A520?
For example:
ASRock A520M-ITX/ac
Supports AMD AM4 Socket Ryzen™ 3000, 3000 G-Series, 4000 G-Series, 5000 and 5000 G-Series Desktop Processors; Supports DDR4 4733+ (OC); 1 x PCIe 3.0 x16; Graphics Output: DisplayPort, HDMI; 7.1 CH HD Audio (Realtek ALC887/897 Audio Codec); 4 x SATA3, 1 x Ultra M.2 (PCIe Gen3 x4 & SATA3); 6 x USB...
If form factor doesn't matter I'm curious: which advantage does a X300 motherboard have compared to e.g. A520?
For example:
ASRock A520M-ITX/ac
Supports AMD AM4 Socket Ryzen™ 3000, 3000 G-Series, 4000 G-Series, 5000 and 5000 G-Series Desktop Processors; Supports DDR4 4733+ (OC); 1 x PCIe 3.0 x16; Graphics Output: DisplayPort, HDMI; 7.1 CH HD Audio (Realtek ALC887/897 Audio Codec); 4 x SATA3, 1 x Ultra M.2 (PCIe Gen3 x4 & SATA3); 6 x USB...www.asrock.com
Doesn't X300 allow for CPU/iGPU overclocking ?
Others in this thread care more about the form factor, as well as the single 19V input rail. For me to get non-APU CPU support on the A300W/X300W Deskmini, need an X300/A300 board that supports those CPUs- it doesn't matter to me if the board was EATX. Hopefully I'll be able to transplant that CPU support into the A300W/X300W UEFI Firmware.If form factor doesn't matter I'm curious: which advantage does a X300 motherboard have compared to e.g. A520?
For example:
ASRock A520M-ITX/ac
Supports AMD AM4 Socket Ryzen™ 3000, 3000 G-Series, 4000 G-Series, 5000 and 5000 G-Series Desktop Processors; Supports DDR4 4733+ (OC); 1 x PCIe 3.0 x16; Graphics Output: DisplayPort, HDMI; 7.1 CH HD Audio (Realtek ALC887/897 Audio Codec); 4 x SATA3, 1 x Ultra M.2 (PCIe Gen3 x4 & SATA3); 6 x USB...
Yes, however there are hardware differences between the X300 and A300- I'm running an A300W with the X300W bios and I still can't overclock.
Thank you, ill try to ask ETA PRIME, im not asking for OC capability, but if we can put 5000G to A300, it will give far better performance upgrade.
@Danlopez1222 Hello, nice to see you being active again! If you're still on track with the EEPROM-Subject.
I've found a reason, why your traces did not provide a satisfying result:
smallformfactor.net
I've managed to extract possible (supported) EEPROM chips, which are used in the UEFI-FW in the version 3.60S.
(Using both, the UEFITool and Ghidra - C++ Reverse of the UEFI-Modules, which are written in C/C++)
There should be a conflict, since you're decrypting/settings the stream for another IC, which is not used by the UEFI-FW.
It may be useful for us - the supported ICs and afterwards the analysis of that particular IC's data as context.
Greetings
I've found a reason, why your traces did not provide a satisfying result:
Prebuilt - [SFFn] ASRock's DeskMini A300 - Finally!
Conclusion: Need to raise SOC VID. Not possible. If SOC VID can be adjusted, the system could run @3600MHz without any problems. Just think for yourself. It ran with XMP Timings on 3200,3333 and 3400 with SOC VID = 0,9 (~1.0V) Renoir is amaizing. @gustav I am a little confused by your post...
smallformfactor.net
I've managed to extract possible (supported) EEPROM chips, which are used in the UEFI-FW in the version 3.60S.
(Using both, the UEFITool and Ghidra - C++ Reverse of the UEFI-Modules, which are written in C/C++)
There should be a conflict, since you're decrypting/settings the stream for another IC, which is not used by the UEFI-FW.
It may be useful for us - the supported ICs and afterwards the analysis of that particular IC's data as context.
Greetings
@Danlopez1222 about the idea of CPU-microcode/-support transplantation:
I think we are a step further to be able to achieve that. There is a product announced: "DeskMini Max" which is 5000-Series based
mainboard with PCIe 16x slot. Therefore, we could probably grab it's BIOS/UEFI and do some surgery on it combined with A300 UEFI
DeskMini Max is using *no* dedicated chipset. Still running on knoll-based-config.
It seems from the video (they're opening up the case, so you can see the mainboard) they've updated the power-stages on the mainboard being
4+1+1 (I guess '4' to be able to satisfy 105Watt against 65W using '3' VRMs)
liliputing.com
www.asrock.com
This platform seems to be indeed the A300 v2.0 because of the PSU, but also TDP being up to 105 Watt it seems they've updated the VRMs and
power-rails for the CPU/SoC/APU - there is residing my hope of "true" o/c being actually feasible. There is only one M.2 slot for NVMe's though and 2x S-ATA.
And being 10L in volume, it remains a question weather a standard mITX/ITX-platform would be actually the "way-to-go"-platform, since 10L is not small factor
anymore - when comparing against 1.92L from the A300. I think the price will be the point, on which a customer can decide.
PSU + Case + Mainboard (B550-based) => ~300€
Deskmini Max therefore should be <300€
And again, you can re-buy the RAM -.-, since Max is using U-DIMMs. Not pleased by this fact.
I think we are a step further to be able to achieve that. There is a product announced: "DeskMini Max" which is 5000-Series based
mainboard with PCIe 16x slot. Therefore, we could probably grab it's BIOS/UEFI and do some surgery on it combined with A300 UEFI
DeskMini Max is using *no* dedicated chipset. Still running on knoll-based-config.
It seems from the video (they're opening up the case, so you can see the mainboard) they've updated the power-stages on the mainboard being
4+1+1 (I guess '4' to be able to satisfy 105Watt against 65W using '3' VRMs)
ASRock DeskMini Max is a bigger small PC with support for Ryzen 5000 and discrete graphics - Liliputing
ASRock DeskMini Max is a bigger small PC with support for Ryzen 5000 and discrete graphics
liliputing.com
ASRock > News
This platform seems to be indeed the A300 v2.0 because of the PSU, but also TDP being up to 105 Watt it seems they've updated the VRMs and
power-rails for the CPU/SoC/APU - there is residing my hope of "true" o/c being actually feasible. There is only one M.2 slot for NVMe's though and 2x S-ATA.
And being 10L in volume, it remains a question weather a standard mITX/ITX-platform would be actually the "way-to-go"-platform, since 10L is not small factor
anymore - when comparing against 1.92L from the A300. I think the price will be the point, on which a customer can decide.
PSU + Case + Mainboard (B550-based) => ~300€
Deskmini Max therefore should be <300€
And again, you can re-buy the RAM -.-, since Max is using U-DIMMs. Not pleased by this fact.
Last edited:
What's the issue with the U-DIMM ? It's just regular desktop DIMMs. Unless it's from an upgrade from "old" Deskmini to Deskmini MAX perspective (and the need for new RAM).
Yea, you totally got the point. It's about the upgrade-perspective.
I do prefer regular desktop DIMMs, but having an A300 pushed me towards SO-DIMMs.
One question, how come you have more "Reactions" to your posts than the actual number of posts?! nice nice hehe
shhh .... that's a secret
EDIT: YAY, 3 more "reaction points" \o/ ... gimme the likes
Last edited:
So I've sold A300 and bought X300, but still with 4650G. Anyone has checked it with 5xxxG - mostly interested in S3 sleep mode Thanks.
Thanks.
Answer from Asrock support
Hi Tomas,
Thank you for choosing ASRock and contacting us.
Regarding your question, model DeskMini X300 Series with Cezanne CPU doesn't support S3.
Have a nice day,
Thanks,
ASRock TSD
Hi Tomas,
Thank you for choosing ASRock and contacting us.
Regarding your question, model DeskMini X300 Series with Cezanne CPU doesn't support S3.
Have a nice day,
Thanks,
ASRock TSD
Hello all,
I have a 5700G running in my X300 with bios 1.70. My summary about the 5700G against the 4750G: It is not worth the change. The X300 limits both apu's due to its limited cooling capability. The additional performance in synthetic benchmarks is little to nothing. In my opinion the 5600G is the best choice as it is not limited and the integrated graphics is as fast as the one from the 5700G. It is the perfect match regarding price and performance.
Best regards
The_Bavarian
I have a 5700G running in my X300 with bios 1.70. My summary about the 5700G against the 4750G: It is not worth the change. The X300 limits both apu's due to its limited cooling capability. The additional performance in synthetic benchmarks is little to nothing. In my opinion the 5600G is the best choice as it is not limited and the integrated graphics is as fast as the one from the 5700G. It is the perfect match regarding price and performance.
Best regards
The_Bavarian
Curiosity
Too busy figuring out if I can to think if I shoul
Platinum Supporter
Bronze Supporter
M...M...M...M...Multi-Tier...Subscriber...
- Apr 30, 2016
- 724
- 838
I was back to this topic the other day...I can manually enter frequencies for the CPU and GPU, but they don't actually do anything. It looks like the chip ignores anything out of spec.
Also, my logic analyzer came in and works great. I was able to tap into my hardware SPI flasher's interaction while reading the bios:
I'm waiting for some more stuff from amazon this afternoon, then I can decode the interaction between the activator chip and the CPU.
MOSI transfer 0x03 seems to be the command Read Data, followed by the address. The FF seems to be the answer. (Empty blocks)
After reversing UEFI and reading up the datasheet for the Fudai flash model , see page 16 of the datasheet - see for yourself, 2 clicks away.
On the X300 the particular address could be a "feature" with it's options in the BIOS, which would then unlock (as an example)
Or by now already used addresses could return other data (values) which would adjust the BIOS - but there must be an interaction happening.
If we somehow can desolder the chip or perform a hardwired read-out by uninstalling the CPU and RAM and using SPI Flasher/Reader, we may be able to grab it's flash content. Seems to be no rocket science. (Could we use Dan's transfer MITM to reverse a flash able .hex with the intercepted data at addresses in game?)
Edit: for AT26DF041 its also 0x03 (see page 10) for low freq. read-command. But do not get distracted by the flash type. If desoldered, by sending 0x9F to the flash it will say who it is. (JEDEC specs, which I did not know was applying for the SPI Flash)
So we would be able to mimic A300 with 3.xx BIOS or (when flash reprogrammed) the X300 with 1.xx BIOS... At least that is the idea...
Edit 2: since more and more people are owners of the X300 model, may I ask for the possibility to take a picture of their knoll chip and the area surrounding this? It would clarify whether on this important stage the boards have any differences. It would be a "nice to have" - info, so we could grasp a better picture.
Last edited:
Similar threads
