Search results

X670 is, literally, two daisy-chained B650, all going through a single PCIe 4.0 x4 link (Intel is now using PCIe 4.0 x8 for its consumers chipsets). And X870 is a minor design update on X670—so minor that there's no urgency to offer the single chip as B850. This looks like a pitiful marketing...

Because the large L3 cache specifically helps with audio, or because it's all new and shiny? Physical x16 slot for GPU (x8 electrical should be enough, if need be) 2 x1 slots for audio cards x4-8 slot for 10G NIC 1 SATA is not a constraint yet some of the M.2 slots will have to come from the...

Which one? Anyway, if you detail your need (cores, clock, number and nature of drives, number and nature of PCIe slots…) we can help you find suitable motherboards while avoiding cable spaghetti as much as possible. That would be: UAD PCIe x1 (?) RME PCI = ? (I'm not in audio; is that "PCI" as...

An onboard NIC needs lanes just like an add-on card… Use adapters to convert these M.2 slots into PCIe slots. Or, if you're really out of lanes, go for Xeon Scalable or Threadripper rather than consumer CPUs.

It's not that difficult to add 10 GbE networking to a Hackintosh. Do not wait for it to be on-board.

In addition to Intel 500 series, with Mieze's kext, and Aquantia 107/113 with CaseySJ's patches, you can use Chelsio T520 cards or Solarflare SFN5122F / SFN6122F cards (but not the "newer" PCIe 3.0 7122F) with the kext driver from the manufacturer (Catalina package still available from Chelsio...

At first sight, I see nothing which could explain the breakage. $ diff F67d_Sinkclose_fix/dsl/dsdt.dsl F67b_no_Sinkclose_fix/dsl/dsdt.dsl 8c8 < * Disassembly of dsdt.dat, Fri Sep 27 15:54:03 2024 --- > * Disassembly of dsdt.dat, Fri Sep 27 15:55:13 2024 14c14 < * Checksum 0x45 ---...

And NootedRed is incompatible with WhateverGreen, which the -wegnoigpu argument implies… You really should make up your mind and get rid of incompatible components.

Keep your kidney… (or at least get a Threadripper 7000 or EPYC system for the price)

Ouch!

So the CPU is fine but Ryzen 9000 will require a new version of OpenCore, and possibly of other kexts, to account for its new CPU family 0x1a. Can you provide @Shaneee with the log from his OC 1.0.2 DEBUG above?

Does it fail to boot Linux or Windows?

Thanks, now the last piece: An ACPI dump after BIOS update (OpenCore DEBUG, Misc>SysReport:true) BIOS updates can, and do, change ACPI tables.

Where do you see that "spoofing is not accurate"? Device-id ix 0x73bf and the macOS graphic driver attaches.

To contribute, please do as suggested in post #1 and post ACPI dumps of your board before and after the BIOS update. We do not have your board and "blocked at startup" is NOT sufficient to understand what could be the issue.

This is 6900 XTXH. And, as explained, you need to name a PCI bridge to apply device properties. Please put WEG back and try this SSDT in place of your SSDT-BRG0; it names the bridge and spoofs a 6900 XT.

What do you need to "fix"? A RX 6900 XT should be supported by macOS. If it really is a XTXH, spoofing a regular 6900 XT should be easier and safer than NootRX.

It is compatible with existing AM5 platforms, and X870(E) is apparently the same silicon as X670/B650, so it looks like BIOS updates could be the main issue.

Er… welcome. But you should be aware that EFI folders are motherboard-specific—and USB maps can even be case-specific, depending on which headers are plugged. So you're basically on your own. You may use an existing EFI as a starting point, but only you or someone with the exact same board as...

This is more a question for @Lorys89 , according to his thread on macos86.it. But if I got his explanations right, this should be it: