Contribute
Register

Gigabyte B550 Vision D (Thunderbolt 3) + AMD Ryzen 7 3700X + AMD RX 5600 XT

CaseySJ said:
The most reliable Thunderbolt behavior we have yet seen (and we have seen a lot of them) are on:
  • Gigabyte Z390 Designare with built-in Titan Ridge (Intel)
  • Gigabyte Z490 Vision D with built-in Titan Ridge (Intel)
Of these two, the older Z390 Designare is the most reliable, with the Z490 Vision D being a close second.

Quite a few musicians and audio engineers are using both of these systems with UAD devices. If you look at Thunderbolt 3 Experiences in the Z390 Designare thread, you will see what I mean.
Click to expand...
thanks for all the tips. Z390 Designare ✨
since we're on Ryzen thread, which boards are on your list @CaseySJ ?
 
Last edited:
CaseySJ said:
Not sure I understand. Can you clarify?
Click to expand...
I meant from the AMD boards that you've worked with (B550 Vision D, X570 Creator, X570 Taichi), which one would you recommend.
From the perspectives of usability, stability, etc.
 
centraljava said:
I meant from the AMD boards that you've worked with (B550 Vision D, X570 Creator, X570 Taichi), which one would you recommend.
From the perspectives of usability, stability, etc.
Click to expand...
Here is a summary:
  • ASRock X570 Creator:
    • This is the first AMD AM4 motherboard I purchased. I was drawn to it because of its on-board Titan Ridge controller.
    • However, hot plug never worked because any attempt to add the crucial PCI-Thunderbolt and PCIHotplugCapable device properties resulted in the collapse of the Thunderbolt device tree.
    • Furthermore, the Creator uses a Macronix SPI Flash ROM chip for Thunderbolt firmware. I was able to read this chip only twice in many multiple attempts over several months using Raspberry Pi, Reveltronics REVELPROG-IS, Arcelli, and standard USB CH341a programmers. It was nearly impossible to read the chip, so I never got a chance to flash it.
    • MacOS itself worked very well and very reliably. Rock solid stability. 10GbE port worked well. Eventually, even sleep/wake worked.
    • But I sold this board when the Gigabyte B550 Vision D was released.
  • ASRock X570 Taichi:
    • This was purchased a couple of months after the Creator. It lacks on-board Thunderbolt, but contains a Thunderbolt header.
    • So I wanted to see if Thunderbolt hot plug might work with an add-in-card. I tried (a) Asus ThunderboltEX 3, (b) Gigabyte GC-Titan Ridge, and (c) ASRock Thunderbolt 3 AIC.
    • The only one that worked (i.e. hot plug) was the ASRock Thunderbolt 3 AIC when connected to the on-board header.
    • However, with Big Sur I'm not sure if hot plug is working (will have to test it soon).
      • UPDATE: Thunderbolt devices do not connect either from cold start, warm start, or hot plug. This just reinforces my disdain for ASRock's abysmal ACPI implementation.
    • MacOS on this system also works very well and is rock solid. However, sleep has stopped working. I suspect my NZXT all-in-one liquid cooler (connected to internal USB 2 header) is to blame, but I haven't spent much time on it.
    • I still have this system and keep it up-to-date with Big Sur.
  • Gigabyte B550 Vision D:
    • This is the best AMD AM4 platform I have yet encountered. I am ecstatic to say that the following work:
      • Thunderbolt hot plug with PCI-Thunderbolt and PCIHotplugCapable properties injected through OpenCore with no problem whatsoever. No need to flash firmware for hot plug.
      • Flashing the Thunderbolt firmware is a cinch with the Arcelli 3.3V USB CH341a programmer. Mine has been flashed with modified NVM33 from Designare Z390.
      • Sleep/wake works!
      • It comes with an Intel AX200 WiFi/BT module, which is a Key A/E M.2 module instead of a CNVi module. This means it can be replaced with a Broadcom BCM94360NG Key A/E module, which is what I have done.
        • So I have WiFi/BT fully working without using up a PCIe slot.
      • It also has a PCIe 4.0 NVMe slot in which I have a PCIe 4.0 NVMe SSD (Sabrent Rocket) that is achieving over 4000MB/s.
    • The system experienced a number of reliability problems with the old AMD R9 380X GPU. But I am happy to report that since installing the ASRock Challenger Pro RX 5600 XT, all of those problems have disappeared completely. The system is totally rock solid.
However, as with all Ryzen-based Hackintoshes, we must keep this in mind:
  • Thunderbolt daisy chaining may not be reliable.
  • Complex Thunderbolt devices may not be reliable (case by case basis).
  • Some UAD audio interfaces may exhibit audio stutter.
  • Some Adobe applications may not work.
  • Virtualization software will not work.
 
Last edited:
On the topic of B550 Vision D stability under macOS, eDEX-UI is another app that works reliably now -- after switching to RX 5600 XT GPU. This is a fancy high-tech version of Terminal.
github.com

GitHub - GitSquared/edex-ui: A cross-platform, customizable science fiction terminal emulator with advanced monitoring & touchscreen support.

A cross-platform, customizable science fiction terminal emulator with advanced monitoring & touchscreen support. - GitSquared/edex-ui
github.com github.com
Screen Shot 2021-01-23 at 10.13.24 AM.png
 
Is some library dedicated to AMD from Linux. Could help us?
 
T2 is not present inside iMac 2019.

So for handle H.265 video transcoding, iGPU is needed for accelerate.
important would be the next design for new iMac (mac Intel or not).

if Apple does not support XE graphics, on new board computer, all mac Intel will be obsolete very shorty, but could be upgrade video in the future. I hope so, because all mac Intel aren't obsolete now. (But could be.)
 
Last edited:
@CaseySJ may I ask you in future OC patch guide add to ACPI patches a condition so that it will not apply in Windows? Or is there any other solution to fix ACPI_BIOS_ERROR bsod screen? Cheers!
 
anthera93 said:
@CaseySJ may I ask you in future OC patch guide add to ACPI patches a condition so that it will not apply in Windows? Or is there any other solution to fix ACPI_BIOS_ERROR bsod screen? Cheers!
Click to expand...
Certainly; working on it now.
 
CaseySJ said:
This is a tricky problem on this AMD Ryzen system. All of the SSDTs are already being injected conditionally, but Windows still fails to boot though OpenCore.

However, let me see if I can make the conditional statements more aggressive.
Click to expand...
Just fixed it for myself. SSDT-X570-Vega64-slot-1.aml was the one crashing windows boot. (that's the one i use, other graphics card fixes may crash too)

Code: 
DefinitionBlock ("", "SSDT", 1, "RLJ", "X570VEGA", 0x00000000)
{
    External (_SB_.PCI0.GPP8, DeviceObj)
    External (_SB_.PCI0.GPP8.D012, DeviceObj)
    External (_SB_.PNLF.BSET, MethodObj)    // 1 Arguments
    External (BRTL, IntObj)
    External (DTGP, MethodObj)    // 5 Arguments
    External (EGPS, IntObj)

    If (_OSI ("Darwin")) // Added this condition
    {
        Scope (_SB.PCI0.GPP8)
        {
            Scope (D012)
            {
                Name (_STA, Zero)  // _STA: Status
            }
            
................
 
You must log in or register to reply here.
Copyright © 2010 - 2024 tonymacx86 LLC
Back
Top