Ryzen 7 4700G — Ventura 13.5 & Sonoma Public Beta — Radeon Graphics 16 GB [Renoir] — NRed

[craighazan]

Try this revised version of Edhawks EFI posted in #139.

 

[shingatedigu]

Not working sir...kindly please try to repair my efi file...because it was working fine with ventura

 

[shingatedigu]

And please guide what bios setting should i do

 

[shingatedigu]

please help me to install macOS Sonoma..I tried to install it with my EFI which I used for Ventura.. it's perfectly working with Ventura but now I am trying to install Sonoma with the same EFI files. Initially, all was good but after completing the installation after the Apple logo it restarted, and when I rebooted again it showed some logs and then shutdown every time..please help me, bro. I am attaching my EFI file and error screenshot...please kindly help me

 

[shingatedigu]

 

[shingatedigu]

Please guys Help me

 

[craighazan]

Please guys Help me

Would you share with us the changes you made to BIOS.
In other news, the NootedRed.kext is working with Sequoia, see link below. But there's a 'But!' in the way, the developers are waiting for lilu to be updated for the Constants.

macOS Sequoia support is ready, but... · ChefKissInc NootedRed · Discussion #275

I'm still waiting for Acidanthera to add the Sequoia constants to Lilu. So I can't actually add them in properly. For the meantime, have some screenshots.

github.com

 

[craighazan]

And please guide what bios setting should i do

Check this page at the bottom. https://dortania.github.io/OpenCore-Install-Guide/AMD/zen.html#starting-point

 

[craighazan]

@Edhawk some exciting news from ChefKiss today!. ExtremeXT is again involved with development of NootedRed and I assume NootRX. Apparently the new TSC kext, ForgedInvariant, and other developments encouraged him to return, I’m very happy about this.

macOS Sequoia support is ready, but... · ChefKissInc NootedRed · Discussion #275

I'm still waiting for Acidanthera to add the Sequoia constants to Lilu. So I can't actually add them in properly. For the meantime, have some screenshots.

github.com

 

[craighazan]

I updated my build to Apples latest macOS, pronounced Suh-Koi-Ahh!. And ooof, what a screen saver!, up there with Montereys in my opinion, Sonoma and Ventura were of the mark I thought. Anyway 'Can I install macOS Sequoia on my APU?', yes you can by following these simple steps.

• Update your AMD Vanilla Patches. Shaneee updated the patches for Sequoia and now contain 25 entries, you still only need to edit the first four to match your core count.
• 

  GitHub - AMD-OSX/AMD_Vanilla at beta

  Native AMD macOS via OpenCore. Contribute to AMD-OSX/AMD_Vanilla development by creating an account on GitHub.

  github.com

Update the following kexts.

• AppleALCU 1.9.0 > 1.9.1 Link
• lilu 1.6.7 > 1.6.8 Link
• RestrictEvents 1.1.3 > 1.1.4 Link
• VirtualSMC 1.3.2 > 1.3.3 Link
• NootedRed Link

As of writing Intel BT isn't working, But Intel Wi-Fi is with Heliport and Itlwm.

• Heliport 1.5.0-alpha Link
• Itlwm 2.3.0 Link

I download the public beta using gibMacOS and double clicked the InstallAssistant.pkg to launch the installer. Despite the 14.45 GB file size, I did update from Sonoma and didn't lose any of my stuff, I suggest installing onto a second HD tbh. For installation I disabled NootedRed, it can take a long time to install and verbose will hang for what seems like an eternity on 'igpu' if the kext is enabled. To avoid the 7 MB VRAM blues, you can enable again after the second restart.

 

[craighazan]

​

It's been a while since I last updated this thread about my 4700G Hackintosh. I use it every day, more so than my M4 mini and my 5800XT Hack!. I wondered why that is, the M4 mini is a Mac, fast, small and does what a Mac should do. The 5800XT is my gaming machine, I rarely switch to macOS unless I want to up date it and the 2TB XG7000 sits at <2% used. But this build is different, I never really thought about it as an AI budget build, it was about NootedRed and gaming, but surprisingly that's what it's ended up being. Of course the 4700Gs advantage, (and other AMD APU's) is it's ability to take RAM and let the APU use that as VRAM, I currently have 32 GB RAM installed and in BIOS I can select UMA SPECIFIED to 16GB to give the APU access to 16 GB VRAM!. That's a lot, you can run large LLMs such as Qwen-Image 2.1 12 GB+ or Flux.2 Klein 9B 10 GB+, though don't expect it to be quick, I downloaded Qwen-Image, a 16 GB+ model and to generate an image took over 10 minutes!, SD 1.5, a much smaller model takes around 5-7 minutes. In addition to the 4700G, I installed a broken Sapphire Pulse RX 6600, (no output), surprisingly it showed up in System Report and it can still be used for AI, though it's limited by its 8 GB VRAM, ToshLLM will let you split tasks between 2 GPUs. But as you can see from the bench below the RX 6600 alone is still the quickest, by some margin. Another positive about this are those listings on eBay for broken AMD 6000 series cards with no output!, they can still be used for compute in AI!.

GitHub - engeldlgado/toshllm: Run large language models locally on Intel Macs with AMD GPUs - native macOS app with Metal acceleration

Run large language models locally on Intel Macs with AMD GPUs - native macOS app with Metal acceleration - engeldlgado/toshllm

github.com