Jump to content

iMac Pro X99 - live the future now with macOS 10.14 Mojave [Successful Build/Successful Guide]


KGP-iMacPro
 Share

107 posts in this topic

Recommended Posts

  • 2 weeks later...

does anyone solved XCPM implementation on an 5820k?

Ive followed most closly the E.1) steps described.

 

Bios settings with latest microcode patch for my Asrock X99m Extreme4,

Bios settings

  • Advanced\CPU Configuration\CPU Power Management Configuration\
  • Enhanced Intel SpeedStep Technology (EIST): disabled)
  • Turbo mode: Enabled
  • CPU C-State: Enabled
  • Enhanced C1 State: Enabled
  • CPU C3 Report: Enabled
  • CPU C6 Report: Enabled
  • Package C State Limit: C6(non Retention) state

a) there is no locked msr register on my asrock x99m extreme4, so no kernelpm and no xcpm core scope msrs pike r patch,

b) added the fakecpu id 0x0306F2

c) active

d) active

e) active

f) also active PluginType in SSDT Patch

 

update 3

now with speedstep, EIST Performance FIX for Broadwell-E and Haswell-E is different:

Broadwell-E

FIND

C1E30848 63D389D0 48C1EA20 B9990100 000F3048 FF05B38F 76004883 C4085B5D C30F1F40 00

 

REPLACE

B800FF00 004863D3 89D048C1 EA20B999 0100000F 3048FF05 B38F7600 4883C408 5B5DC390 90

 

Haswell-E

FIND

C1E30848 63D389D0 48C1EA20 B9990100 000F3048 FF05938D 76004883 C4085B5D C30F1F40 00

 

Replace

BB00FF00 004863D3 89D048C1 EA20B999 0100000F 3048FF05 93AE7600 4883C408 5B5DC390 90

 

Still I dont think im done with XCPM Performance

Geekbench 4 Results with 5820k at 4GHz low, and intel power gadget cpu frequency curve is low:

779594130_Bildschirmfoto2018-12-06um18_48_45.thumb.png.47c903d950f6e43fbf763121579f4d4c.png

1705538016_Bildschirmfoto2018-12-06um18_49_14.png.e4d3595e6e6db65a827fdbf5fc51c028.png

 

 

P-States etc:

607738175_Bildschirmfoto2018-12-08um10_15_47.thumb.png.89268e662ea3ff78e91870699f651953.png

 

 

USB

In addition, can someone help me with my board-specific XHC USB kext for Asrock X99m Extreme4?

 

 

OLD SCREENSHOTS

 

 

 

 

 

 

Bildschirmfoto 2018-12-06 um 18.09.21.png

 

Edited by Multisaft
  • Like 1
Link to comment
Share on other sites

On 12/5/2018 at 11:25 AM, Multisaft said:

does anyone solved XCPM implementation on an 5820k?

Ive followed most closly the E.1) steps described.

 

Bios settings with latest microcode patch for my Asrock X99m Extreme4,

Bios settings

  • Advanced\CPU Configuration\CPU Power Management Configuration\
  • Enhanced Intel SpeedStep Technology (EIST): disabled)
  • Turbo mode: Enabled
  • CPU C-State: Enabled
  • Enhanced C1 State: Enabled
  • CPU C3 Report: Enabled
  • CPU C6 Report: Enabled
  • Package C State Limit: C6(non Retention) state

a) there is no locked msr register on my asrock x99m extreme4, so no kernelpm and no xcpm core scope msrs pike r patch,

b) added the fakecpu id 0x0306F2

c) active

d) active

e) active

f) also active PluginType in SSDT Patch

 

update 3

now with speedstep, EIST Performance FIX for Broadwell-E and Haswell-E is different:

Broadwell-E

FIND

C1E30848 63D389D0 48C1EA20 B9990100 000F3048 FF05B38F 76004883 C4085B5D C30F1F40 00

 

REPLACE

B800FF00 004863D3 89D048C1 EA20B999 0100000F 3048FF05 B38F7600 4883C408 5B5DC390 90

 

Haswell-E

FIND

C1E30848 63D389D0 48C1EA20 B9990100 000F3048 FF05938D 76004883 C4085B5D C30F1F40 00

 

Replace

BB00FF00 004863D3 89D048C1 EA20B999 0100000F 3048FF05 93AE7600 4883C408 5B5DC390 90

 

Still I dont think im done with XCPM Performance

Geekbench 4 Results with 5820k at 4GHz low, and intel power gadget cpu frequency curve is low:

779594130_Bildschirmfoto2018-12-06um18_48_45.thumb.png.47c903d950f6e43fbf763121579f4d4c.png

1705538016_Bildschirmfoto2018-12-06um18_49_14.png.e4d3595e6e6db65a827fdbf5fc51c028.png

 

 

P-States etc:

607738175_Bildschirmfoto2018-12-08um10_15_47.thumb.png.89268e662ea3ff78e91870699f651953.png

 

 

USB

In addition, can someone help me with my board-specific XHC USB kext for Asrock X99m Extreme4?

 

 

OLD SCREENSHOTS

 

 

 

 

 

 

Bildschirmfoto 2018-12-06 um 18.09.21.png

 

 

1.) You have SMT_scope_msrs1 enabled twice. KGP-D1.2_e_xcpm_SMT-scope_msr1 is certainly not the patch of my 10.14 EFI-Folder and is wrong!

 

2.) Also KGP_d1.2_10.13_xcpm_pkg_scope and KGP_d1.2_10.13_xcpm_pkg_scope are not from my 10.14 EFi-Folder and are wrong! The correct patch is xcpm_pkg-scope_msrs, which you disabled!  

 

3.) Up to my knowledge there is just one common performance patch valid for both Broadwell-E and Haswell-E. 

 

The correct performance patch for 10.14.2 is:

 

Find: C1E3084863D389D048C1EA20B9990100000F3048FF05C37276004883C4085B5DC30F1F4000
Repl: B800FF00004863D389D048C1EA20B9990100000F3048FF05C37276004883C4085B5DC39090

 

4.) You do not need the cupid_set_info kernel patch. 

 

Hope this helps. 

 

Link to comment
Share on other sites

On 12/10/2018 at 7:38 PM, KGP-iMacPro said:

 

1.) You have SMT_scope_msrs1 enabled twice. KGP-D1.2_e_xcpm_SMT-scope_msr1 is certainly not the patch of my 10.14 EFI-Folder and is wrong!

 

2.) Also KGP_d1.2_10.13_xcpm_pkg_scope and KGP_d1.2_10.13_xcpm_pkg_scope are not from my 10.14 EFi-Folder and are wrong! The correct patch is xcpm_pkg-scope_msrs, which you disabled!  

 

3.) Up to my knowledge there is just one common performance patch valid for both Broadwell-E and Haswell-E. 

 

The correct performance patch for 10.14.2 is:

 


Find: C1E3084863D389D048C1EA20B9990100000F3048FF05C37276004883C4085B5DC30F1F4000
Repl: B800FF00004863D389D048C1EA20B9990100000F3048FF05C37276004883C4085B5DC39090

 

4.) You do not need the cupid_set_info kernel patch. 

 

Hope this helps. 

 

 

Hi KGP,

thanks for your feedback. Ive gone through your recommendations ones more and disabled my kerneltopatch entries which I found for haswell-E in various topics (insanelymac.com, most of them from nmano): 

 

with your recommended patches i dont get xcpm speedsteping on 10.14.1, this is the bootstraß patch, the performance patch and smt scope 1/2 patches you recommend in your tutorial.

1338554350_Bildschirmfoto2018-12-15um16_09_51.thumb.png.4336199812512a1bf68621fe7a0b5e4a.png

 

 

my geekbench score is at the level I would say is max. for 4gh 5820k with slow 32gb ddr4 2133 mhz ram, but the cpu does not idle anymore, all the time full 4ghz frequency.

1735717491_Bildschirmfoto2018-12-15um16_12_50.png.fb14f0e721e2e7df86f960bfc0e683e7.png

1669113580_Bildschirmfoto2018-12-15um16_14_53.thumb.png.069bb3e4b39416183cb7f7d970a5ff4a.png

Link to comment
Share on other sites

7 minutes ago, Multisaft said:

 

Hi KGP,

thanks for your feedback. Ive gone through your recommendations ones more and disabled my kerneltopatch entries which I found for haswell-E in various topics (insanelymac.com, most of them from nmano): 

 

with your recommended patches i dont get xcpm speedsteping on 10.14.1, this is the bootstraß patch, the performance patch and smt scope 1/2 patches you recommend in your tutorial.

1338554350_Bildschirmfoto2018-12-15um16_09_51.thumb.png.4336199812512a1bf68621fe7a0b5e4a.png

 

 

my geekbench score is at the level I would say is max. for 4gh 5820k with slow 32gb ddr4 2133 mhz ram, but the cpu does not idle anymore, all the time full 4ghz frequency.

1735717491_Bildschirmfoto2018-12-15um16_12_50.png.fb14f0e721e2e7df86f960bfc0e683e7.png

imageproxy.php?img=&key=c17464ef1b1b51c7

1669113580_Bildschirmfoto2018-12-15um16_14_53.thumb.png.069bb3e4b39416183cb7f7d970a5ff4a.png

 

Because you have to change the bootstrap patch to Haswell-E! The one implemented in my EFI-Folder distrubution is for Broadwell-E! I told you that the correct bootstrap patch for Haswell-E  and further guidelines for the 5820K you can find in the README.md of my X99 EFI-Folder repository! 

Edited by KGP-iMacPro
Link to comment
Share on other sites

1 hour ago, KGP-iMacPro said:

 

Because you have to change the bootstrap patch to Haswell-E! The one implemented in my EFI-Folder distrubution is for Broadwell-E! I told you that the correct bootstrap patch for Haswell-E  and further guidelines for the 5820K you can find in the README.md of my X99 EFI-Folder repository! 

 

Hi KGP 

ok next round after update to 10.14.2 with fakecpuid 0x0306F2

 

1) cpuid_set_info disabled

2) xcpm_bootstrap haswell-e enabled (from your readme.md on x99 repository

3) xcpm_performance_patch for 10.14.2 as posted above enabled

4) smt_scope_msrs 1/2 enabled

5) xcpm_pkg_scope_msrs enabled

 

no with xcpm mode 1 but just base frequency. the cpu does push to higher frequencies after boot:
1928270066_Bildschirmfoto2018-12-15um17_28_25.thumb.png.2081b23456369ac896f6fa68f38ca789.png

Link to comment
Share on other sites

Just now, Multisaft said:

 

Hi KGP 

ok next round after update to 10.14.2 with fakecpuid 0x0306F2

 

1) cpuid_set_info disabled

2) xcpm_bootstrap haswell-e enabled (from your readme.md on x99 repository

3) xcpm_performance_patch for 10.14.2 as posted above enabled

4) smt_scope_msrs 1/2 enabled

5) xcpm_pkg_scope_msrs enabled

 

no with xcpm mode 1 but just base frequency. the cpu does push to higher frequencies after boot:
1928270066_Bildschirmfoto2018-12-15um17_28_25.thumb.png.2081b23456369ac896f6fa68f38ca789.png

 

But the CPU goes above 4 GHz.. Show IPG snapshot for CPU IDLE and CPU max. Load (2-3 cinebench runs) conditions! 

Link to comment
Share on other sites

18 minutes ago, KGP-iMacPro said:

 

But the CPU goes above 4 GHz.. Show IPG snapshot for CPU IDLE and CPU max. Load (2-3 cinebench runs) conditions! 

yes right at startup the cpu goes to 4ghz and than idles back as all the chrome windows etc. are loaded.

sorry what is an IPG snapshot?

 

cinebench does not wake the cpu up:
1032908315_Bildschirmfoto2018-12-15um17_50_15.thumb.png.93ba895972fa049e616a05f3775e7d5e.png

 

appleintelinfo does not find pstates of 40 etc.:
1329596964_Bildschirmfoto2018-12-15um17_51_43.png.3a51d9acb335dcb7774585a211560ce9.png

 

also switched to a sapphire pulse rx580 8GB, but no displayport out, only hdmi 30hz 4k (i know i need to follow your instructions on that).

CPU is unkown:

2076680564_Bildschirmfoto2018-12-15um17_54_08.png.31ed59b4b1d384be517a14966daac1c5.png

 

Edited by Multisaft
Link to comment
Share on other sites

1 minute ago, Multisaft said:

yes right at startup the cpu goes to 4ghz and than idles back as all the chrome windows etc. are loaded.

cinebench does not wake the cpu up:
1032908315_Bildschirmfoto2018-12-15um17_50_15.thumb.png.93ba895972fa049e616a05f3775e7d5e.png

 

appleintelinfo does not find pstates of 40 etc.:
1329596964_Bildschirmfoto2018-12-15um17_51_43.png.3a51d9acb335dcb7774585a211560ce9.png

 

Plugintype checked in config.plist? TSCAdjustReset.kext in EFI-Folder? 

 

Revise BIOS settings and play enabling, disabling the xcpm patches. 

 

You do have different board but also CPU. I cannot help you further. However, there are several people successfully running the 5820K based on my guidelines. 

 

Good luck,

 

KGP 

Link to comment
Share on other sites

18 hours ago, KGP-iMacPro said:

 

Plugintype checked in config.plist? TSCAdjustReset.kext in EFI-Folder? 

 

Revise BIOS settings and play enabling, disabling the xcpm patches. 

 

You do have different board but also CPU. I cannot help you further. However, there are several people successfully running the 5820K based on my guidelines. 

  

Good luck,

 

 KGP 

 

Yes Plugintype is checked (activated), TSCAdjustReset.kext is patched to 11 in accordance to your guide (12-1).  I always had problems with the X99M under el capitan, maybe its time to switch.

 

I've bought an used X99-A II, I will start from scratch when it arrives.

Edited by Multisaft
Link to comment
Share on other sites

Hi KGP,i have same problem like Multisaft

 

i use your guide since 10.12.

my hardware

2696v3(haswell ep)

asrock x99 taichi

xcpm is loaded,pstate ok(12-38).bios patched.

 

update from 10.13.2 to 10.14.2

with your guide.

all cpu core will lock to 1.2Ghz after pc run some hours.

 

if i disable xcpm bootstrap patch,xcpm won't load,all core lock to 2.8Ghz.

 

if i use xcpm performance patch and xcpm msr patch in this post

 

the cpu freq will vary from 1.2 to 2.8,even i force run some heaveload program(cb r15),there may some core still keep 1.2-2.0.this make cb r15 score vary from 1000-2000(normal is 1900-2100) points in multi test.

make me feel like i'm using a 18 cores low freq notebook...

 

i think haswell-ep need another eist performance patch or xcpm msr patch.but these info is too less.

does anyone else with haswell-ep has cpu freq issue?

 

i'm not use this pc now,if u have time,need some info to help,i'll post info late.

 

 

 

Link to comment
Share on other sites

1 hour ago, KGP-iMacPro said:

 

I clearly outline the differences between Broadwell and Haswell  in README.md in my X99 EFI-Folder Github repository. 

 

thanks your reply.i read your guide and git readme carefuly.

 

I'm sure i set the right xcpm_bootstrap,FakeCPUID to haswell ep version(2696v3),disable or not the set_cpu_info patch not influence these.

and TSCAdjustRest,bios microcode patch,etc... set to right.

 

in  nmano's post,

he use 

C1E3084863D389D048C1EA20B9990100000F3048FF05238376004883C4085B5DC30F1F4000
BB00FF00004863D389D048C1EA20B9990100000F3048FF0593AE76004883C4085B5DC39090

yours
C1E3084863D389D048C1EA20B9990100000F3048FF05C37276004883C4085B5DC30F1F4000
B800FF00004863D389D048C1EA20B9990100000F3048FF05C37276004883C4085B5DC39090

 

the _xcpm_pkg_scope_msrs _xcpm_SMT_scope_msrs 1 & 2 patch is different.

 

your patch will cause my cpu lock to 1.2Ghz after some hours.

his patch will cause my cpu freq vary from 1.2-2(half cores),2-2.8(half cores),random,which cause cinebench r15 score vary from 1000-2000.(normal 1900-2100).IGP package freq vary from 1.x to 2.x(below 2.5 even high load...)

 

in your guide post in another forum,i noticed someone(id:torock) use xeon 2660v3 also has same problem(xcpm ok,but freq not stable).

 

I guess some inaccurate kernel to patch entities cause these.and there's no much info about these thing.

now my pc like a 18 cores notebook with low freq,not fluence.the performance even worst than my i7 8700(mojave).

I consider change cpu to xeon 1680v3|v4.i use recc memory.i need xeon.

 

thanks your nice guide again.

Link to comment
Share on other sites

23 minutes ago, itsme001 said:

 

thanks your reply.i read your guide and git readme carefuly.

 

I'm sure i set the right xcpm_bootstrap,FakeCPUID to haswell ep version(2696v3),disable or not the set_cpu_info patch not influence these.

and TSCAdjustRest,bios microcode patch,etc... set to right.

 

in  nmano's post,

he use 

C1E3084863D389D048C1EA20B9990100000F3048FF05238376004883C4085B5DC30F1F4000
BB00FF00004863D389D048C1EA20B9990100000F3048FF0593AE76004883C4085B5DC39090

yours
C1E3084863D389D048C1EA20B9990100000F3048FF05C37276004883C4085B5DC30F1F4000
B800FF00004863D389D048C1EA20B9990100000F3048FF05C37276004883C4085B5DC39090

 

the _xcpm_pkg_scope_msrs _xcpm_SMT_scope_msrs 1 & 2 patch is different.

 

your patch will cause my cpu lock to 1.2Ghz after some hours.

his patch will cause my cpu freq vary from 1.2-2(half cores),2-2.8(half cores),random,which cause cinebench r15 score vary from 1000-2000.(normal 1900-2100).IGP package freq vary from 1.x to 2.x(below 2.5 even high load...)

 

in your guide post in another forum,i noticed someone(id:torock) use xeon 2660v3 also has same problem(xcpm ok,but freq not stable).

 

I guess some inaccurate kernel to patch entities cause these.and there's no much info about these thing.

now my pc like a 18 cores notebook with low freq,not fluence.the performance even worst than my i7 8700(mojave).

I consider change cpu to xeon 1680v3|v4.i use recc memory.i need xeon.

 

thanks your nice guide again.

 

Well my patches are exactly those of @PMheart. The SMT and pkg_scope patches are different between 10.13 and 10.14. The performance patch nearly changes for each macOS version. Are you sure you are using the correct patches for your macOS version? Between Broadwell and Haswell there is just a difference in the cpu_set_info and bootstrap patch. While both are obsolete for Broadwell-E, some Haswell CPU s still need the bootstrap patch. Else your CPU behavior will also strongly depend on your BIOS settings. If you have locked MSR, you also need to enable the core scope patch.

 

I guess you already considered all these advices. Unfortunately, I am not able to provide further help in this case. 

 

With my i7-6950X everything works flawless with the patches implemented in my respective EFI-Folders. I can absolutely exclude that some of the patches do not meet the requirements.  

 

Good luck,

 

KGP  

 

 

Edited by KGP-iMacPro
Link to comment
Share on other sites

52 minutes ago, KGP-iMacPro said:

 

Well my patches are exactly those of @PMheart. The SMT and pkg_scope patches are different between 10.13 and 10.14. The performance patch nearly changes for each macOS version. Are you sure you are using the correct patches for your macOS version? Between Broadwell and Haswell there is just a difference in the cpu_set_info and bootstrap patch. While both are obsolete for Broadwell-E, some Haswell CPU s still need the bootstrap patch. Else your CPU behavior will also strongly depend on your BIOS settings. If you have locked MSR, you also need to enable the core scope patch.

 

I guess you already considered all these advices. Unfortunately, I am not able to provide further help in this case. 

 

With my i7-6950X everything works flawless with the patches implemented in my respective EFI-Folders. I can absolutely exclude that some of the patches do not meet the requirements.  

 

Good luck,

 

KGP  

 

 

 

thanks for your reply.

 

i use the right version performance patch for 10.14.2.follow your guide carefully.

asrock x99 taichi,has no msr lock.so i don't need core scope patch and clover kernelpm config.

you confirm your kernel patch i'll try your guide sometime(may next version)

 

i search the web find some guys(three-five use xeon v3) have similar performance problem(xcpm loaded,cpu freq not stable,lower bench score).later if u known someone use xeon 26xx v3 suceess without performance problem.could you post it in your guide.then i know it's some fault in my config.i'll check my setup even consider a clean install.

 

thanks for your help.

Edited by itsme001
Link to comment
Share on other sites

Well thanks to this aweson guide well prepared and the TIME consume of this elite member, after a while my self been inactive on the Hackintosh hobbit, I got a succesed results first shot with this guide, it can not be no more precise of what it is already. Thanks a lot KGP for your kindness.

 

 

 

GB3-3925-61953

SS-1

SS-2

SS-3

SS-4

Link to comment
Share on other sites

 

Well on GB-4 on Mojave I got a pretty preatty good result I think, in comparison the one I got in Sierra withe the same system setup. I never test this this system with High Sierra cause I was out of the hobbit for a while, but this are the results for Mojave.

 

post-1404459-0-51451200-1480780857_thumb.png 

 

SIERRA

 

GB4-4472-5139

 

Mojave

 

Edited by The Veteran
Link to comment
Share on other sites

  • 2 weeks later...

Hi Guys,

 

short feedback,

I've updated to the X99-A II and it looks like XCPM with 5820k cpu work flawless. sadly the sapphire pulse rx580 does not work the way I expected it. Just HDMI does work and my Philipps 40" does not support 4k at 60Hz over HDMI (old version), and all the Displayport 1.2 solutions do not work as well (yes I used the RX580.aml). So Sleep/wake issues are also present (black screen problem). I decided to invest even further as I need this workstation for both (WIN as well as Mac) and ordered a PowerColor Radeon RX Vega 56 Red Dragon. I hope that this will fix my Displayport as well as sleep/wake issues. 

 

I just want to thank KGP for his support and all the other guys for such incredible easy to install hackintoshes. I'm around in this community for a long time (not as active as some others) and did my first x86 on an Asus P6T-SE with Core i7-920 and holy this was a messup to get OS X running. What a difference it is today when you just stick to hardware used from our community!

Link to comment
Share on other sites

ok

 

now with powercolor vega56 red dragon

 

and without displayport 1.2 connection to my 40", so no 4k@60hz. and not even 4k@30hz.

with hdmi, 4k@30hz is possible for my tries:

 

1. try: just added your vega56 ssdt implementation, this resulted in massive fanspeeds due to your oc and fan control settings, I guess. with or without whatevergreen.kext no displayport connection.

2. try: deleted everythin, your ssdt implementation, now still no displayport connection, idle fan control works with 0rpm in idle but under load normal vega problems with high fan speeds and no reversion back to low fan speeds or zero fan speed.

 

so I googled and found the various informations in tonymacs about VGTab control.

there, I understood, that I dont have my Vega56 on GFX0, I searched in my config.plist that I used from your folder KGP and found that the change GFX0 to IGPU was enabled in the dsdt patching section. I disabled it, but still there is no GFX0 found from gfxutil (also from tonymacs support sites).

 

Now I'm lost.

 

Next idea: update lilu and whatevergreen in recommendation from the VGTab guide

 

Link to comment
Share on other sites

12 minutes ago, Multisaft said:

ok

 

now with powercolor vega56 red dragon

 

and without displayport 1.2 connection to my 40", so no 4k@60hz. and not even 4k@30hz.

with hdmi, 4k@30hz is possible for my tries:

 

1. try: just added your vega56 ssdt implementation, this resulted in massive fanspeeds due to your oc and fan control settings, I guess. with or without whatevergreen.kext no displayport connection.

2. try: deleted everythin, your ssdt implementation, now still no displayport connection, idle fan control works with 0rpm in idle but under load normal vega problems with high fan speeds and no reversion back to low fan speeds or zero fan speed.

 

so I googled and found the various informations in tonymacs about VGTab control.

there, I understood, that I dont have my Vega56 on GFX0, I searched in my config.plist that I used from your folder KGP and found that the change GFX0 to IGPU was enabled in the dsdt patching section. I disabled it, but still there is no GFX0 found from gfxutil (also from tonymacs support sites).

 

Now I'm lost.

 

Next idea: update lilu and whatevergreen in recommendation from the VGTab guide

 

 

I really do not understand at all what you are doing.. Vega56 SSDT has no loadtable. No OC! 

 

Everybody runs Vega56 SSDT + Whatevergreen without any issues including fully functional display ports! 

 

I clearly recommend to use VGTab app for the application of custom vega settings by creating a custom load table to be inplemented within the SSDT! 

 

In my config plist there is nothing enabled in the graphics section and there is certainly no change from GFX0 to IGPU implemented or enabled!

 

Seems that you already  totally messed up my predefined EFI-Folder. I really would recommend to start from scratch in your case. 

 

Good luck

Edited by KGP-iMacPro
Link to comment
Share on other sites

On 1/10/2019 at 6:31 PM, KGP-iMacPro said:

 

I really do not understand at all what you are doing.. Vega56 SSDT has no loadtable. No OC! 

 

Everybody runs Vega56 SSDT + Whatevergreen without any issues including fully functional display ports! 

 

I clearly recommend to use VGTab app for the application of custom vega settings by creating a custom load table to be inplemented within the SSDT! 

 

In my config plist there is nothing enabled in the graphics section and there is certainly no change from GFX0 to IGPU implemented or enabled!

 

Seems that you already  totally messed up my predefined EFI-Folder. I really would recommend to start from scratch in your case. 

 

Good luck

 

good morning KGP,

 

I really do not understand what I'm doing different to others, too! :-D

But after my last post I did what I written down, updated to the latest Lilu, without any Vega56 SSDT but with Whatevergreen.kext and now I'm on 4k@60Hz and with GFX0. As I had run out of time yesterday, I will follow your SSDT Vega 56 implementation today and will retest yours as well as creating my own (undervolting and Idle fan speeds etc. - just as I like to do it on my own as it interests me). What I read in the VGTab tutorial, there are users reporting issues using the kext created from VGTab in combination with the whatevergreen.kext (failed boot procedures). Is that the reason you implemented the various correct OC and fan-settings for vega in your ssdt implementation?

 

I will give you a screenshot of the config plist I downloaded (with date of download), it is the config I downloaded from your first post in this thread and which I used to clearly had no problem creating a USB installer as well as taking the first steps in the clean Mojave install. I don't think that I totally messed up my EFI-Folder, I just updated in accordance to your guide. I can upload my config.plist if you like to or if you want to recheck if I failed in other steps :-D

 

To be honest I have absolutely no problem with the fact that I run into some minor issues, that helps me to understand what we are doing with our installations. 

 

Regarding the Vega56 SSDT, can you tell me if you undervolted P5-P7 and how I can change your SSDT to 0rpm fanspeed at idle?

 

Benchmarks

Update: when I use your Vega56 SSDT with GFX0, there is no 0rpm fanspeed at idle. Can you give me a hint how to fix this? I see 1.61GHz GPU clocks as well as 900MHz GPU Ram speeds. that is nice, looks like you undervolted?

754457862_Bildschirmfoto2019-01-11um18_20_33.thumb.png.4610063f396e8954ce2938d0f982b95c.png

 

In addition, I found out that my pkg Power delivery is not monitored in intel power gadget. See attached my config.plist and screenshots.

1836298329_Bildschirmfoto2019-01-11um18_19_31.thumb.png.8064e5208388ad560994e02183b1ee03.png

 

711332083_Bildschirmfoto2019-01-11um18_20_52.png.8419f10698fd121d77b3a6894ba849c0.png

 

Luxmark gives me even 1.65GHz at GPU speed, nice.

 

1946240719_Bildschirmfoto2019-01-11um18_41_22.png.cd4fda6906636d52868f1fd4df0d5112.png

1149732332_Bildschirmfoto2019-01-11um18_42_22.png.cab0caedb1256af219bd4cc92a83b2c8.png

1010799040_Bildschirmfoto2019-01-11um18_45_26.png.5e40c5d9e4b5485cf45d6b1360d776a7.png

 

CONFIG PLIST

config.plist

 

 

here my vega56 SSDT

SSDT-X99-Vega56.aml

undervolted and oced to 1.66ghz with 900mhz ram clock, lowest possible idle rpm (0rpm does not work?)

 

results almost the same as above

 

 

 

Edited by Multisaft
Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
 Share

×
×
  • Create New...