fabiosun

https://postimg.cc/image/4hmjdtxw7/"> Ha scritto dopo 43 secondi: entrambi sbagliati :) pero' senza nessun kext accessorio

vecchi driver con il kext WEG (Whatevergreen di vit ) adatti i vecchi per avere un supporto minimo (dual display e risoluzioni) ma senza accelera∑ione Ha scritto dopo 16 secondi:

https://postimg.cc/image/i1eb5dkiv/"> tutto ok WEG ha funzionato e le USB ci sono Ha scritto dopo 1 minuto 45 secondi: https://postimg.cc/image/6qbngny4n/">

evviva :) provo as vedere se la trova semplicemente in App Store :)

Use other software to check (Istat menu IE) and test with cinebench single core or Geekbench or cinebench graphics test with all of this you will use less cores and frequency should bounce higher

Welcome :)

Ciao @obus IPG graphics are pretty similar You should do also 3 test if you can Geekbench (CPU and GPU) in free version only OpenCL is available and it is ok then in cinebench do some test also for gpu (as I was asking before) and for cpu single core like you can see in my attachment I am asking because I can't see in both IPG graphics higher freq than 3.3 or 3.4 I think In some condition you should see some peak around 4.0 and more however outpinfo show all states so your cpu scales well in both case (few missing states are irrelevant..they could come far in time during testing Ha scritto dopo 7 secondi: Ha scritto dopo 1 minuto 9 secondi: then important other things..stability is the same? temperature seems pretty identical in both cases Ha scritto dopo 9 minuti 58 secondi: Ha scritto dopo 4 minuti : during single CPU and Graphic cinebench run

run 4 times Cinebenchz for cpu and graphic with appleintelinfo terminal executed also during test run Intel power gadgets and post here image of these datas for both fakecpuid togheter complete appleintelinfo output (you have to run command in termal many times (sorry if you know this) you have to test for about 15/20 minutes...

load and idle load and idle normal usage...

ideal condition CPU C6-Cores [ 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 ] CPU P-States [ 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 (33) 34 35 36 37 38 39 40 41 42 43 ]

check your PM

se stai avviando da usb provane un'altra

yep another input https://ark.intel.com/products/130046/Intel-Xeon-W-2175-Processor-19_25M-Cache-2_50-GHz">https://ark.intel.com/products/130046/I ... e-2_50-GHz">https://ark.intel.com/products/130046/Intel-Xeon-W-2175-Processor-19_25M-Cache-2_50-GHz see your CPU specs turbo for few cores is @4300 so CPU P states must be 43 and you have to see all in CPU PSTATES your cpu has 14 cores / 28 Threads this means you have to see in CPU C6 states all from 0 to 27.. when you see this situation without any hole..you are gold (technically speaking) :) in Italian way is simple to figure out this technically speaking..I don't know in English if it is the same :)

C3 states could be visible only programming MSR 0xe2 from this value 0x1E000005 to another with the help of a kernel patch which an user )SammlerG) found for old OSX version If I was you I would test with software like cinebench, Geekbench, luxmark 3.1 all your system and post the result also grabbing output and temperature with IPG To see performances, stability and temperatures in both condition I think however which your system could reach all States also with Skylake x fakecpuid..maybe it needs more time to reach all states then you should have a clear situation of optimal condition for your system Ha scritto dopo 5 minuti 41 secondi: to explain better C3 states stuff (this is an extraction from an old Insanely Mac thread I created time ago) follow patch I was using in Sierra to have C3 States, in bold the one which does what I have said in previous message (MSR 0xE2 programming) Solution 2 (same results and behavior of solution 1) Bios Unlocked EIST Enabled ​4g enabled (only for Sierra) ​kext Patch IOPCIFamily find 4881FB00000040 replace 4881FB00000080 5960X 10.12 Brumbaer 6 patches find AA010000DC3300000000000000000000000000000000000001 replace AA010000000000000000000000000000000000000000000000 SammlerG Sierra xcpm MSR Patch 1 / xcpm support Pike R. Alpha find A0010000DC33000000000000000000000000000040000000010005 replace A0010000DC33000000000000000000000000000040000000010004 SammlerG Sierra xcpm MSR Patch 2 / xcpm support Pike R. Alpha find 3A060000DC33000000000000000000001F00000000000000000000000000000000000000000000000000000000000000 replace 3A0600000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 SammlerG Sierra xcpm MSR Patch 3 / xcpm support Pike R. Alpha find 42060000DC33000000000000000000001F0000000000000018 replace 42060000000000000000000000000000000000000000000000 SammlerG Sierra xcpm MSR Patch 4 / xcpm support Pike R. Alpha find E20000004C00000000000000000000000F040000000000000500001E replace E20000004C00000000000000000000000F040000000000000700007E SammlerG Sierra xcpm MSR Patch 5 / xcpm support Pike R. Alpha [/color] find 89D8C1E008B99901 replace B800250000B99901 [ I use 25 (3700 Mhz) you can change accordly your OC clock] (wrsmr fix to get CPU max) © okrasit You can disable a couple of them if needed ;-)

Ciao obus It seems we can not avoid using fakecpuid.. Probably also others like Broadwell Ep 0x40764 (I don't remember well) do the same job It is very strange also you can use Appleintelinfo kext because in new I9 architecture it is not working well if MSR is checked in option Analyzing it you miss C3 states in output but it is pretty normal Previously when I had an unsupported cpu there was a kernel patches by SammlerG to activate C3 output, in real use however no difference to come back to our "issue" I am searching a kernel patch for adding cpu id in kernel section..but I can't find it Also it would be the same thing of using fakecpuid, so no great catch No I do not thing it is a clover bug? but also it is strange if you use KernelXCPM flag is not working for you.. this flag was an automated way to patch unsupported Cpu we had asked the past to clover's coders to not use in automatic way. My advice is to launch appleintelinfo and use your system loaded/unloade for few time and see if all CPU pSTATES come out then evaluate if for you it is better to use a Broadwell EP fakecpuid or Skilake X in terms of stability and performances Hoping Apple programmers include your cpu in next OSX release Ha scritto dopo 1 minuto 51 secondi: PS remember also to try unlock MSR bios option to see if you can avoid using MSR patch or any kernel patches Using a fakecpuid you are using a way to use XCPM of another family of CPU But it not should be your case because Xeon W family is A CPU supported in iMac Pro...( ;( )

Sorry @obus your first MSR patch togheter fakecpuid produces 1 for both terminal command?

ciao ti chiedevo anche i software in quanto per la scheda grafica ancora oggi e' possibile scegliere Nvidia o AMD anche se il supporto apple e' orientato su AMD

uhm for me today is enough if you can try boot strap togheter fakecpuid..and test if XCPM and Frequency Vectors are set to 1 good evening :) Ha scritto dopo 40 secondi: tomorrow we will try other solution :)

ok in that condition could you flag in config KernelXCPM and see

If you have a dual system with also old High Sierra you can use both usb patches specifying matchOS accordly have you test latest config?

https://www.gigabyte.com/FileUpload/global/MicroSite/473/x299.html Strumento interessante per il posizionamento delle periferiche (Gpu/Nvme) per sfruttare al massimo le "lanes" della propria cpu per possessori di gigabyte X299

interesting tools to configure properly lanes cpu/GPU/Nvme To achieve optimal performance for pciexpress slot for X299 Gigabyte owners https://www.gigabyte.com/FileUpload/global/MicroSite/473/x299.html

for 10.14.x AppleUSBXHCI find 83FB0F0F 83030500 00 Replace 83FB0F90 90909090 90 MatchOS 10.14.x sorry I am concentrated on CPU..I haven't seen you are not using correct one for Mojave Ha scritto dopo 5 minuti 52 secondi: also you can find here (in Italian language) a summary of changement USB port limit thread

It is the one you use in your config.plist use it togheter boot strap and see (without fakecpuid obviusly) :) Ha scritto dopo 1 minuto 35 secondi: if possible put also here actual efi you are using in the tries Thanks Ha scritto dopo 5 minuti 53 secondi: obus, if you can download from here this tool https://www.macos86.it/MacOS86_Downloads/Applicazioni/Info/Hack_Info.zip">MacOS86_Downloads/Applicazioni/Info/Hack_Info.zip">https://www.macos86.it/MacOS86_Downloads/Applicazioni/Info/Hack_Info.zip post zip it produce. in this way is it possible to check other aspect