Jump to content
TheBloke

Five or six simultaneous displays with AMD GPU (eg 7970)?

5 posts in this topic

Recommended Posts

Hi all

 

I currently use an NVidia 760 with four displays: 1 x 4K; 2 x 1920x1200; 1 x 1920x1080.   This mostly works, but I am considering replacing it with an AMD GPU - likely a Radeon 7970 - for two reasons:

  1. I would like to add a fifth monitor, maybe even a sixth;
  2. I get small stutters and slowdowns with my NVidia, using both Native and NVidiaWeb drivers.  It is a lot better than it was with my NVidia 980Ti, but it's still not perfectly smooth as it should be, especially when I have a full screen video playing while also doing UI movements like Swipe Left A Space or Mission Control.

 

I have just been told that it should be possible to access all display outputs on an AMD 79XX GPU, possibly requiring a custom SSDT and/or radeonDeInit?  I've only ever had one AMD GPU and that was 8 years ago so I am not experienced with them.  My NVidia 980Ti has five outputs, but can only use four at once.  But I believe AMD's EyeFinity does allow it to use them all as separate displays.

 

Before I buy the GPU, I'd be most grateful if anyone could confirm that it should be possible to run five or six simultaneous displays on a card such as the ASUS HD 7970 DirectCU II (6 x DP outputs) or XFX AMD Radeon HD 7970 (2 x mini-DP; 1 x HDMI; 2 x DVI).  With one display being 4K @ 60fps and the rest 1920x1200 or 1080P.

 

I have Googled on this a while but haven't yet found anyone discussing connecting more than four monitors to an AMD Hack  I did see that the late-2013 Mac Pro (which uses 2 x AMD FirePro GPUs) states it can support "up to 6 Thunderbolt displays".  So I know the OS can do it, it's just a question of whether a particular GPU can, and in a Hack setup.

 

Thanks very much

 

TB

Share this post


Link to post
Share on other sites
Advertisement

I think I have this figured out now:

  • Yes, AMD Eyefinity allows for a full six independent displays and it works on macOS
  • Normally this is achieved using DisplayPort Multi-Stream Technology (MST), which involves hubs or adapters to split one DP port into multiple display connections
  • But the Asus 7970 card I listed is able to support a full six connections without adapters.
  • However, EyeFinity only allows for a maximum of two "legacy" connections.  Legacy means VGA, DVI or HDMI.
  • Only two of my monitors support DP, so that means with my current screens I could only run four at once: 2 x DP + 1 x DVI + 1 x DVI -> HDMI.  Which is exactly what I already have.
  • Using normal DisplayPort to HDMI/DVI cables or adapters would not help - this is still considered a DVI or HDMI connection, streamed within the DP protocol.
  • However, there is an Active DisplayPort to HDMI adapter which converts DP to HDMI and is considered by the GPU to be true DP to DP .
  • That adapter specifically states it works with EyeFinity, and that a maximum of two adapters can be used.
  • Therefore, I could achieve six displays with the following configuration:
    • DP -> DP
    • DP -> DP
    • DVI -> DVI
    • DVI -> DVI
    • DP -> active adapter -> HDMI
    • DP -> active adapter -> HDMI
  • In my case I would likely only use five displays, so I would only need one active adapter.  I do actually have a sixth screen, but it's another DVI.  There's also an active DP -> DVI adapter from a different manufacturer, but I really don't have room for six screens so I don't think I will need to go down that road :)  (If I did want to, it seems that I could use a standard HDMI to DVI passive cable/adapter plugged into the HDMI port.)

 

I am pretty sure that is all correct.   I'm sure enough to order a used Asus 7970, anyway. 

 

Of course it took me just long enough to work this out that the card I planned to Buy it Now on eBay has sold.  But that may not be the worst thing, because I've realised there's another Asus 7970 which is slightly faster, and there's one on auction that ends on Tuesday.   As well as another DirectCU II that ends Monday.

 

Thanks for listening ;)

Share this post


Link to post
Share on other sites

I'm just going to leave this here..

IRNJIO8.png

LFjNoUDl.jpg

:lol:

Seven displays baby.    Well, OK, six displays connected to the GPU.  And an iPad connected via Duet.  Still, seven!

This is achieved with the Asus Matrix 7970 3GB I mentioned earlier.  It's not yet perfect - after boot I only get a picture on two displays; the others have a signal and show up in Displays, but only show black.  The fix is to sleep and then wake.  Which only works with WhateverGreen installed, otherwise wake fails and I have to reboot.  But with WEG added I can sleep from the Login screen and immediately wake to get a picture on all screens, and then I'm good to go until reboot - including plugging/unplugging/swapping any connection/cable.

I don't plan to run with all seven displays.  The one on the far right I added just to check I could, and for purposes of the photo :) I will almost certainly lose that one, and go with 'just' the other five + iPad.  Also I need to move the TV on the left because I can only see 60% of it from where I sit; the rest is blocked by the two screens to its right.  I'm going to mount it a little higher and a little further away so I can see it all.

And finally I need to continue trying to fix the black-screen-after-boot issue.  I've spent many hours trying to patch a FrameBuffer, as per VladA's guide and using InjectATI=Yes in Clover.  No luck so far - all results are worse than what I get booting without any Clover graphical config at all.  But I will continue trying.  Or, an option I've yet to try is using a custom SSDT to try and achieve the same thing.

But I'm pretty happy that I can get the displays working at all, and even if I never solve the on-boot issue, having to sleep&wake after every reboot is certainly not the end of the world.

 

Edited by TheBloke

Share this post


Link to post
Share on other sites

very cool.  I have a FirePro W9000 connected to 3 Dell UP2715K's (2 DP cables per monitor).  On startup, getting all the monitors to sync up is somewhat random ... often, one screen will be blank.  I'm not sure what triggers this.  Less monitors connected seems to make the problem less bad, however.  I've been wondering if it's a signaling or voltage issue with all monitors connected.

 

WhateverGreen is, I think, more essential for proper operation than people realize.  It may seem like issues have gone away now with 10.13.4 and eGPU support, but I'm not convinced that is the case, based on all of the other fixes it seems to make.  Apple is really bad about hardcoding all the settings into the OS configuration ... often, improperly.

Share this post


Link to post
Share on other sites

Just dont get a RX 480, I never managed to run quad display configuration with it, two of them always stay undetected (blackscreen) , even with WhatEverGreen.

 

It seems older cards like 7970 can handle it well.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Recently Browsing   0 members

    No registered users viewing this page.

  • Similar Content

    • By Fede1132
      Hi guys i have a problem install macOs Mojave on my PC, I have this configuration:

      CPU: Ryzen 7 2700X;
      GPU1: GTX 970 Strix(For game rendering and 144Hz Monitor);
      GPU2: GTX 750ti Strix(For 2nd monitor rendering to get 100% performances from 1st 144Hz Monitor);
      MotherBoard: MSI x470 GAMING PLUS
      Hard Disk 1 (Windows 10): Samsung EVO M.2;
      Hard Disk 2 (For Storage): RAID (1) between two hard disks;
      Hard Disk 3 (For Mojave): Kingston SSD 120GB;

      I have created my USB with this:
       
      Create Bootable USB Code: sudo /Applications/Install\ macOS\ Mojave.app/Contents/Resources/createinstallmedia --volume /Volumes/USB -- /Applications/Install\ macOS\ Mojave.app -- nointeract && say Done. Clover Settings: Boot: Verbose; dart=0; nv_disable=0; kext-dev-mode=1; GraphicsEnabler=No; XMPDetection = No. Cpu: C6. Devices: (Not Enabled = [ ] - Enabled = [-] - Full Enabled [✓]) Inject (Full Enabled); Add ClockID (Full Enabled); FixOwnership (Full Enabled); Gui: (Not Enabled = [ ] - Enabled = [-] - Full Enabled [✓]) Enabled (Full Enabled); Screen Resolution: 1920x1080; Kernel and Kext Patches: (Not Enabled = [ ] - Enabled = [-] - Full Enabled [✓]) Apple RTC (Full Enabled); KernelPm (Full Enabled); SMBIOS: iMac(18,3); System Parameters: Inject Kexts = Yes; Inject System ID; Used Kexts: FakeSMC FakePCIID NVIDIA WEB LILU BIOS: VT-d = Disabled; XHCI Hand-Off = Enabled; Legacy USB Support: Enabled; Hard Disk: AHCI Mode; Selection Mode: Legacy + EUFI; Clover Install Log:
      Full Error Log:
       
    • By corbrink
      Hi,
      I have the following system:
      - Gigabyte Z370M-D3H M-ATX
      - Core i7-8700K Coffee Lake
      - Gigabyte Radeon RX 560 4GB
      - Crucial Ballistix 16GB DDR4
      - 960 EVO 1TB NVMe SSD
      - 850 EVO 500GB SSD
      - Crucial 500GB SSD
      - The 2 500GB SSD's run in Raid for data storage.
      - Western Digital 3TB HDD - Time machine
      - EFI here (too large to attach): EFI.zip

      Questions:
      1. I have a dual screen running from Radeon card, 1 on DP and 1 on HDMI. The one screen on DP goes randomly blank from time to time. Any ideas?
      2. I've read that the kexts should preferably be installed in /library/extensions. I would like to know what kexts should remain in the EFI. I read that you should install FakeSMC in both locations (EFI and L/E). If I follow this route must I change something in the config.plist?
      3. USB info in System Information (attached does not seem correct. I've followed the changes suggested but I'm not sure if this is the best it can be.
      4. Are there something in the EFI and config.plist I don't need or doing wrong?

      Feedback will be appreciated.


    • By 1Ale1
      Ciao a tutti, ultimamente ho deciso di mettermi al lavoro e fare un investimento a lungo termine comprando un fisso che ho assemblato pezzo dopo pezzo leggendo numerosi forum per essere bene informato sul miglior hardware che potessi ottenere per un Hackintosh ottimizzando al meglio i costi: ho finito così per comprare:
       
      Hardware:
       
       
      -Motherboard: Gigabyte Z390 Aorus Pro (Rev 1.0)
      -CPU: Intel i7 9700k raffreddato a liquido (CoffeLake)
      -GPU: AMD Radeon RX580 (avendo letto le eccellenti prestazioni a ottimo prezzo per MacOS)
      -RAM: 32GB DDR4 Crucial Ballistix Sport LT (4x8GB 3000MHz)
      -NVMe SSD Samsung 970 Evo 250GB M.2 dedicato solo a MacOS, nel secondo slot della Mobo M.2 avrei poi messo un Western Digital Blue da 250GB con Windows 10 su metà disco e per l'altra metà il Debian che sto utilizzando adesso
       
       
      Poi mi sono messo subito al lavoro per cercare di far partire questo benedetto MacOs su una macchina non ufficiale, ho tovato quindi questa guida su GitHub che era molto simile alla mia configurazione:
       
      https://github.com/cmer/gigabyte-z390-aorus-master-hackintosh/blob/master/STEP_BY_STEP.md
       
      Dopo aver fatto partire una macchina virtuale per configurare la USB come comunemente viene fatto, scaricando l'applicazione e facendo il "burn/copia dei file" di macOS su pennetta esterna rendendola bootabile solamente da un pc ufficiale mac, sono passato alla configurazione di Clover (l'EFI installer, cartella che allegherò qui sotto, ma prima leggete il problema): Ho messo i driver necessari, mettendo quelli che lui menzionava sotto la sezione "building the usb installer"
       
      https://hackintosh.gitbook.io/-r-hackintosh-vanilla-desktop-guide/building-the-usb-installer
       
       
      Install Clover for UEFI booting only
      Install Clover to the ESP
      Under Drivers64UEFI:
      AptioMemoryFix (the new hotness that includes NVRAM fixes, as well as better memory management) (che poi ho dovuto rimuovere, leggi sezione https://github.com/cmer/gigabyte-z390-aorus-master-hackintosh/blob/master/STEP_BY_STEP.md#fixing-kernel-panics-at-rebootshutdown)
      VBoxHfs-64.efi (or HFSPlus.efi if available) - one of these is required for Clover to see and boot HFS+ volumes. If you see the option to enable it in the installer, make sure it's selected - if you don't see it in the installer, verify that one of them exists in the EFI -> CLOVER -> drivers64UEFI folder
      ApfsDriverLoader - (Available in Dids' Clover builds - or here) this allows Clover to see and boot from APFS volumes by loading apfs.efi from ApfsContainer located on block device (if using AptioMemoryFix as well, requires R21 or newer)
       
      applicando piccole modifiche a causa di un errore già riportato da questa persona che ha scritto la guida di GitHub di sostituire AptioMemoryFix-64 (dalla cartella EFI/EFI/CLOVER/driver64UEFI) con OsxAptioFix2Drv-free2000, questa infatti riportava  quanto segue dopo aver detto di aver sostituito AptioMemoryFix-64.efi con OsxaptioFix2Drv-free2000:
       
      "I am told that there are downsides (that I don't fully understand yet) to using OsxAptioFix2Drv-free2000, so do this at your own risk."
       
      Ho configurato poi il config.plist
       
      Come diceva lui, partendo da questo file: https://hackintosh.gitbook.io/-r-hackintosh-vanilla-desktop-guide/config.plist-per-hardware/coffee-lake e applicando le modifiche che sotto ha descritto, del tipo:
       
      Open coffeelake_sample_config.plist with Clover Configurator (right click → Open With → Clover Configurator)   In SMBIOS: - Click the button with an up/down arrow (middle right). Chose iMac18,3. This is important since we'll be connecting our monitor to the RX580. The HDMI port on our motherboard is NOT yet working for Hackintoshes. - Make sure the serial number generated is an iMac (mid-2017) by clicking Model Lookup. - Ensure that Check Coverage reports that the serial is NOT valid. You don't want to use somebody else's serial number. - While you're here, copy your Board Serial Number to your clipboard. You'll need it soon.
      In Rt Variables: Paste your Board Serial Number in the MLB field. Set CsrActiveConfig to 0x0 which enables SIP for extra security. This should work just fine for a Vanilla Hackintosh install and is how genuine Macs ship. In Boot: Change the Custom Flags to: shikigva=40 uia_exclude=HS14 (this disables onboard Bluetooth since we'll be using an external Broadcom Wi-Fi/Bluetooth adapter) In ACPI: Click List of Patches and enable the following: Change GFX0 to IGPU In Devices: Set Inject to 16. Now to enable our headless iGPU, we need to fake the device id. To do so, Click Properties, select PciRoot(0x0)/Pci(0x2,0x0). Then, click the + button to add a property. Add the following: Property Key: device-id Property Value: 923E0000 Value Type: DATA Click the Export Configuration button (bottom left), then Save As config.plist. Copy your newly generated config.plist to /EFI/CLOVER/ on your bootable USB key.  
      Impostazioni BIOS
       
      E avendo settato nel BIOS le impostazioni che questa perona configliava all'inizio della guida,                       (N.B. la mia versione di Bios è l'ultima, F7 https://www.gigabyte.com/Motherboard/Z390-AORUS-PRO-rev-10#support-dl-bios) (a quanto risulta il giorno 7.01.2019)
       
      Load Optimized Default Settings Peripherals → USB Configuration → XHCI Hand-off : Enabled Chipset → Internal Graphics : Enabled (important for Quicklook/Preview) Please note that we will be using our internal GPU in headless mode only and this guide assumes that. This is how an iMac18,3 (what we're basing our build on) behaves. That's it! I literally didn't change anything else and it just worked.
       
      Anche per me ha funzionato. Comunque lui riportava ulteriori settaggi, aggiungendo:
       
      However, these are settings that are generally recommended. Your mileage may vary:
       
      BIOS → Fast Boot : Disabled BIOS → LAN PXE Boot Option ROM : Disabled BIOS → Storage Boot Option Control : UEFI Peripherals → Trusted Computing → Security Device Support : Disable Peripherals → Network Stack Configuration → Network Stack : Disabled Peripherals → USB Configuration → Legacy USB Support : Auto Peripherals → USB Configuration → XHCI Hand-off : Enabled (Extremely important) () Chipset → Vt-d : Disabled (non l'ho messo e parte a volte, quindi penso che se parte non ha una importanza così rilevante) Chipset → Wake on LAN Enable : Disabled (era già così) Chipset → IOAPIC 24-119 Entries : Enabled  
       
      Il sistema è partito e soltanto alcune volte dava il seguente errore:
       
      Error allocating 0x11996pages at 0x000000001d453000 alloc type 2
      Couldn't allocate runtime area.
       
      Credo che sia questo piccolo problema: la NVRAM.
      Non riesco bene a capire dove cercare per sapere con esattezza se è supportato nativamente o meno. Ho cercato di informarmi, voi potreste dirmi se è vero?
      Premendo poi Ctrl+Alt+Canc riavviavo e selezionando da Clover l'NVMe interno partiva senza dare questo errore, per cui il kernel non trova spazio perché la memoria Non Volatile ad accesso Randomico è troppo frammentata, come
       
      Questa guida riportava:
       
      Primo post:
      https://nickwoodhams.com/x99-hackintosh-solving-osxaptiofixdrv-allocaterelocblock-can-not-allocate-relocation-block/
       
      Post aggiornato nel 2016:
      https://nickwoodhams.com/x99-hackintosh-osxaptiofixdrv-allocaterelocblock-error-update/
       
      Mi sono letto tutto e ho capito che lui applicava anche delle modifiche al file: "Changes to OsxLowMemFixDrv-64.efi were simply editing 0x10000000 to 0x20000000."
       
      Ho quindi seguito il suo consiglio di mettere OsxAptiofix Emuvariable 64 e Test2-2016.efi (che lui aveva creato) ma il problema sembrava persistere, più raramente
       
      La mia cartella driver64UEFI di ora
       
      Dopo aver trovato una strada senza soluzoni anche lì mi sono rivolto a degli amici, e di questo errore ne ho parlato anche con utenti del gruppo Telegram "Hackintosh Italia" che sono stati e ancora sono molto gentili ad aiutarmi spiegandomi che si tratta di un problema di NVRAM, che nel mio caso, avendo installato questi .efi file nella cartella EFI/EFI/CLOVER/driver64UEFI (che ancora è cosi attualmente, oscillando tra couldnt allocate runtime area (anche dopo aver definitivamente staccato la corrente dopo aver spento il PC per capire se poteva aiutare) e a volta anche nessun errore)
       
       
      -ApfsDriverLoader-64.efi
      -DataHubDxe-64.efi
      -EmuVariableUefi-64.efi
      -FSInject-64.efi
      -HFSPlus-64.efi
      -NTFS-64.efi
      -NvmExpressDxe-64.efi
      -OsxAptioFix2Drv-free2000.efi
      -PartitionDxe-64.efi
      -UsbKbDxe-64.efi
       
      In più ho aggiunto gli RC Scripts (come mi è stato consigliato dai membri del gruppo telegram) sul disco interno in cui mac è installato.
       
      Togliendo quindi qualsiasi test.efi o test2- 2016.efi ed ascoltando chi ne sapeva di più. Ora mi trovo a dover affrontare diverse volte questo errore dovendo riavviare per risolverlo, cosa che mi fa dedurre che Osxaptiofix2 2000 sia un po' instabile nel pulire la NVRAM prima di avviare la macchina/ allo spegnimento di questa.
      Questo problema di couldnt allocate runtime area persiste sebbene prema F12 dalla schermata di Boot di Clover.
       
      Vi chiedo gentilmente una mano, inoltre mi è stato detto che siccome il sistema da:
       
      -"3.6 GHz unknown nella sezione "Processor"
      -non riconosce la RAM con un Vendor ID? (del tipo Ballistix dovrebbe essere?)
       
       
      Serve di rivedere il config.plist che ho creato, rivedere i DSDT e gli efi? sono a posto?, (mi hanno detto che gengik84 fare, ma accolgo volentieri l'aiuto di tutti) ,  vi mando una foto della schermata di "About this Mac" e della EFI in allegato. Quello che farò sicuramente mentre attendo vostre risposte è il downgrade da 10.14.3 Beta a 10.14.2, l'upgrade l'ho fatto pensando che magari avrei aggiustato qualcosa, ma l'errore come lo dava prima lo dà ora, non credo che questo cambi molto le cose in gioco. Comunque è meglio non essere in Beta per aggiustare gli errori.
       
       
      Grazie mille per essere arrivato a leggere fino a questo punto, lo apprezzo molto.
       
       
      Per la cartella EFI: sono 22MB zippati, in allegato non entrano...
       
      https://drive.google.com/open?id=106jXkDB0Id045lzIIzAn9F5MwnzWFv3X
       

       
       
       
       
       
       
       
       
       
       
       
       
       
       
       
       
       
       
       
       

       
       
       
       



    • By valueimagery
      Hello Everybody,

      First Hack here, and I have a fresh vanilla install with Mojave 10.14.2 closely following this vanilla installation:
      https://hackintosh.gitbook.io/-r-ha...p-guide/config.plist-per-hardware/coffee-lake

      I was surprised how well it worked initially with the above suggested settings. Only a few minor things (USB SSDT, signing the Bluetooth Kext, and switching from VritualSMC to FakeSMC) needed to be adjusted to get were I am currently.

      I built this machine to be a video editing/photography workstation. And so far pretty happy but would really appreciate if anyone has any knowledge or solutions with the few problems I'm still experiencing:

      1) FIXED (SEE COMMENT BELOW)
      Sleep issues waking up from sleep (maybe due to the ASMEDIA USB controller)
      -FIX WAS UNRELATED TO ASMEDIA CONTROLLER BUT DUE TO PREFERENCE OF "ALLOW BLUETOOTH DEVICE TO WAKE THIS COMPUTER"/ AN EXTERNAL USB 3 to SATA Cable for Hot plugging SSD's/2.5" HDD's.  With USB to SATA adapter unplugged/ Bluetooth preference changed, Mac now sleeps uninterrupted.

      2)
      The ASMEDIA USB 3.1 Gen 2 ports ejecting after sleep

      3)
      Am I currently using the correct setup for graphic acceleration with a connector-less ID for the UHD 630 and Radeon RX580? Specifically for optimized performance with FCPX. 

      4)
      Normal Power Consumption?


      I've attached my config.plist, ioreg.zip from RUNME app, SSDT-UIAC.aml (created using Rehabman Method), and my DSDT.aml (from this post for the Maximus x code courtesy of @MaLd0n ) If anyone has the time to take a look and maybe shed some light if I've done something wrong that be awesome.

      And for anyone looking to build a similar machine, I would highly recommend it, and feel free to use these settings. If these minor last things are addressed, this is a really solid hack.

      Cheers,
      Shark

      Detailed below is What is Not Working, The Build/BIOS Setup/Install Process, and What is Working.


      NOT WORKING:
      1) Sleep:

      If I put the computer to sleep, or it falls asleep on it's own, it wakes up immediately after sleep.
       
      I’m getting these errors in terminal trying to figure this out using the following two commands:

      Command: pmset -g
      sleep 10 (sleep prevented by powerd, sharingd)

      Command: log show --style syslog | fgrep "Wake reason"
      localhost kernel[0]: (AppleACPIPlatform) AppleACPIPlatformPower Wake reason: XHC


      Here are my current settings:
      Clover:
      Boot -> Darkwake=0 (tried 1 and 2 as well)
      Energy Saver Settings: 
      Nothing Checked EXCEPT Put Hard disk to sleep when possible

      Would like Power Nap Enabled for the purpose of Time Machine, but doesn’t seem to be an option with Hackintosh?

      Using Terminal I have disabled the following:
      sudo pmset -a ttyskeepawake 0
      sudo pmset -a proximitywake 0
      sudo pmset -a standby 0
      sudo pmset -a autopoweroff 0

      2) USB 3.1 Gen 2

      For the 2.0/3.0 Intel Chipset, I’ve tried both the @headkaze Intel FB-Patcher v 1.6.8 and @RehabMan instructions on creating a SSDT. The FB-Patcher did not work completely. So I continued with @RehabMan method, created a SSDT, and sacrificed 2 x usb 2.0 ports and 1 internal RGB LED header to reach my 15 port max with 3.0 speeds on all SS ports.

      The ASMEDIA Ports work OOB with full 3.1 speeds. However the Eject issue can only be resolved by using an App like Jettison. And using that app doesn’t solve the sleep issue, computer still wakes. I’d rather figure out why the disks keep ejecting on sleep and if fixing issue #1 would resolve issue #2 or vice versa?

      Do I possibly need a power management ssdt? And how would I create one for the ASMEDIA chip?
       
      I've now tried using the attached DSDT.aml in my Patched folder.  I can now see it's working as my PCI devices in System Profiler are now populating (before it was stating no PCI devices).  However still experiencing sleep and eject issues.

      Ioreg Reads for the ASMEDIA 3.1 controller as:

      The 3.1 controllers for the back USB gen 2 ports are controlled by:
      RP05@1C,4/IOPP/PXSX@0/PXSX@01000000
      PRT1@01100000 (type C) (3.1 speeds)
      PRT2@01200000 (type A) (3.1 speeds)
      PRT3@01300000 (type C) (2.0 speeds)
      PRT4@01400000 (type A) (2.0 speeds)

      The 1 x 3.1 type C port on the front panel is controlled by:
      RP07@1C,6/IOPP/PXSX@0/PXSX@01000000
      HS01@00300000 (2.0 speed)
      SS01@00100000 (3.1 Speed) 

      There is also a SS02@0010000 under this controller, and as far as I can tell it does not have a function. Is there a way to disable, such as when creating an Intel Chipset SSDT?

      3) FINAL CUT ISSUES/ GRAPHIC ACCELERATION:
      Currently it doesn’t appear that the IGPU UHD 630 is doing much in iStats. In Apple System Profile/ Activity Monitor GPU window I only see RX580, no UHD 630. Is this how it should be? Or is that what connector-less ID should look like?

      With FCPX there is no crashing, glitchy exports, or rendering issues, however not sure if Graphic acceleration is being used to its fullest potential.
      The IGPU uses 5-8% Memory/ Processor when playing/scrubbing in timeline.
      And only 5-10% of the IGPU Memory (no processor) is used during any render or export of footage. The RX580 is still doing most of the hard work. My Bruce X times are quick 14-18 seconds, but I’m curious as to why the IGPU isn’t doing any of the lifiting. Is this expected? I'm pretty happy with how Lightroom and Final Cut are performing, I just want to make sure I've set the hardware up correctly.


      4) Power Consumption
      I have a general understanding of building computers, but not an expert especially with first time building a Hackintosh, overclocking, timings, voltages, etc…

      In HWMonitor the CPU Package Core reads blank, CPU Package Total is anywhere from 3.52w to 4.8w, the CPU Package DRAM is anywhere from 1.12 – 5.08. 

      Is this normal, or high? My temperatures seem fine and idle between 45-57 C, And Max shortly at 85 during stress tests.

      So just not sure if the CPU Power Consumption HWmonitor reporting is normal or abnormal. I've attached two HWmonitor screenshots 1 normal idle cpu, one while FCP rendering.
       
      THE INSTALL…
      Kexts Used:
      AirportBrcmFixup.kext
      AppleALC.kext
      BRCMFirmwareData.kext
      BrcmNonPatchRAM2.kext
      BrcmPatchRAM2.kext
      FakeSMC_ACPISensors.kext
      FakeSMC_CPUSensors.kext
      FakeSMC_GPUSensors.kext
      FakeSMC_LPCSensors.kext
      FakeSMC_SMMSensors.kext
      FakeSMC.kext
      IntelMausiEthernet.kext
      Lilu.kext
      USBInjectAll.kext 
      WhateverGreen.kext
      XHCI-unsupported.kext

      BUILD:
      ASUS ROG Maximus X CODE
      i7 8700k OC’d to 4800
      2 x 16gb Corsair Vegeance LPX DDR4 2400mhz C16 OC’d to 2666mhz
      Sapphire RX 580 NITRO+ 8Gb Graphics Card
      BCM94352Z Dell DW1560 (Wifi and Bluetooth Both Natively Working with Rehabman Kexts)
      BeQuiet! Darkrock Pro 4 
      EVGA 650w g1+
      Phanteks Enthoo Evolv X
      Samsung 28” E590DS (DP), Samsung 24” D300 (HDMI), Samsung 22” Syncmaster (DVI), Sony Bravia 48” (HDMI)

      Pre Install BIOS Settings:
      Updated Bios to Version 1704
      Loaded Optimized Defaults 
      AI Overclocker Tuner > Manual
      Extreme Tweeking > Enable
      Advanced \ System Agent(SA) Configuration →IGPU Multi-Monitor : Enabled
      Advanced \ System Agent (SA) Configuration →Vt-d: Disabled
      System Agent (SA) Configuration > Graphics Configuration > Primary Display > PEG
      USB Configuration > Legacy USB Support > AUTO
      USB Configuration > Keyboard and Mouse Simulation > Disabled
      APM Configuration > Power on by PCI - E/PCI > Disabled
      CSM: Disabled 
      Fast Boot > Disabled
      Boot Logo Display > Disabled
      Secure Boot > OS Type > Other OS
      DVMT Pre-Allocated > 128M

      WORKING:
      ICLOUD, iMessages, Handoff, etc

      Nitro+ 8GB GPU:
      4 Displays (DVI, DP, HDMI 1&2) from RX580 with Audio
      Currently set to 1920 resolution, but have option to use up to 4k resolution (3840x2160)
      Audio over DP and HDMI
      *Not sure if Airplay works as I don’t have an Apple TV, but registers in System Bar and say's looking...

      IGPU:
      Set in Connectorless State (AAPL,ig-platform-id: 0300923E)
      *If in a Connected State, slows down Bruce 5k Exports to well over a minute, with glitches, and does not send display signal to DP or HDMI (just green screen), even with connector type patch to correct type. So keeping it connectorless.
      *Not sure if Graphic acceleration is really working as FCPX is struggling with Multicam original media. This could just be the GH5 4k file encoding that FCPX struggles with. Was just hoping to remove creating proxy media from my workflow.

      Audio Front/ Back
      All working:
      Digital Optical Out 
      Back Green - Internal Speakers
      Front Headphones - Internal Speakers automatically changes to Headphones when plugged in
      Back Mic (Internal Microphone, not Line In)
      Front Mic (Internal Microphone, changes to a 2nd Line In)

      Bluetooth
      After initial install Bluetooth was seen by Mac System but not recognising any devices
      Thanks to @RehabMan Entered following code in terminal and bluetooth now works:
      sudo nvram boot-args="kext-dev-mode=1"

      I've now enabled this in my config.plist as well.
      https://github.com/RehabMan/OS-X-BrcmPatchRAM/blob/master/README-Mac.md
      (unsigned kernel extensions need to be enabled. Take the following steps in the Terminal)

      FANS/ TEMPS
      Temps or Fans were not registering with TG Pro using VirtualSMC.kext
      Fixed by switching from VIRTUALSMC.kext with older FAKESMC.kext with sensor gauges kexts. All working now.
      Computer Idles between 46 and 52 Celcius. Hits 81-87 when stress testing.

      USB 2.0/3.0 
      It appears that are physically too many USB ports on the Asus Maximus Code X to ever work with the 15 port limit safely. I created a custom SSDT that disabled USB 2.0 hs07(L),HS08(ML),12(aura led controller) to get usb 3.0 speeds on ss04,ss05, and ss06
      Be curious if anyone ever found a workaround for this...
       
       
      Overclocking:
      RAM:
      2400mhz to 2666mhz with Manual DRAM frequency change in BIOS. Did not change Latency or Timing and no crashing or panics so far. The 2666mhz speed is recognised in About this Mac and the Novabench RAM score increased.
      *Please advise if this method of overclocking the RAM is unadvisable.
       
      CPU:
      4.8ghz @ 75F @ 50% (Prime 95 test)
      Followed these BIOS steps:
       


       
      config_ValuesImac.plist
      DSDT.aml
      Send me Values-iMac.zip
      SSDT-UIAC.aml


    • By daniel_iversen
      Hi there,

      TL;DR; Any idea on how I can get an ATI Radeon HD5570 graphics card working on High Sierra?

      I just installed High Sierra, and I read it was supposed to be “Out of the Box” to get these AMD/ATI cards working, but it’s been very hard and I’ve read like 100 articles and done a million experiments to no avail.

      The card is built in to a HP TouchSmart 610-1030a touch screen PC, and the symptom/problem is:

      - I can at best get a decent resolution at best, but no acceleration (I'm testing with the Chess app but can visually see it everywhere) and the OS only detects 7MB memory whereas it has 1GB.

      I’ve tried a combination of:

      - OOTB (then there’s no detection really)

      - Clover setup in “Graphics” tab - combination of these settings:

      - Inject ATI

      - Load VBIOS/patch VBIOS

      - Entered my card details (what I thought) - vendor ID 0x1002, product ID 0x68d9, FBName “Eulemur” or “Baboon”

      Also;

      - I have Lilu.kext and WhateverGreen.kext

      - I've got the "EmuVariableUefi-64.efi" file deployed in my drivers64UEFI folder 

      - Tried using “downgraded” (from Sierra) versions of kexts AMDRadeonX3000.kext, AMDRadeonX3000GL.Driver.bundle, IOAccelerator2D.plugin, IOAcceleratorFamily2.kext (I think my wifi disappeared with the IO ones which makes sense I suppose)


      At worst in the "About this Mac" it just says "Display 3MB" (and often with poor resolution), and at best it says "ATI Mobility Radeon HD 5570 7 MB" but with no acceleration (and no recognition of the 1GB VRAM).

      Any ideas of what I’m supposed to do? I don’t really get all the stuff I read about patching and rolling my own kexts etc but my impression was that after Sierra I wasn’t supposed to do that for the AMD/ATI cards.

      Thanks a lot in advance!!
×