Contribute
Register

The "Official" C612 (aka lets built the machine Apple never gave us)

Status
Not open for further replies.
@catswift, just a heads up I contacted metacollin about power management. He solved it for 10.12 using multiple kernel patches unique to the generation of CPU's. Should see a guide in the near distant future. :)

@volition awesome, I look forward to reading it! Maybe working xcpm will squeak a few thousand more out of my geekbench score :).

In other news, my old webcam was acting unreliably. Dug into it and it turns out to be a Sierra issue and not a hackintosh issue. Bought a new Logitech c920 which supports UVC (USB Video Device Class) mode and so far working flawlessly. Now I need to look into why connecting to a vpn over l2tp not working...
 
:eek::clap::clap::clap:dude, amazing. :bows head: Any free time to put together a tutorial? I, as I am sure many others, will be eternally grateful. Question, how is software performance? Have you tested any "heavy software" software to see if they hang or lag? I remember reading a few threads on another forum with a successful dual core xeon build but within graphics and video editing software, performance was horrible. Let me know.

I am very pleased to report....

Total success! :D :D :D I found the piece of the puzzle, and it effects all multiprocessor systems and I am fairly confident that this will solve the strange hangs, freezes, and performance issues that users like @analogo have been suffering.

upshot_gwglySKf.png


I now have native xcpm working, complete with all the weird extra (C10 anyone?) Broadwell C-States, in my case, 21 P-States, and single and multicore turboboost working flawlessly. On the Broadwell-EP platform, with dual CPUs.
 
:eek::clap::clap::clap:dude, amazing. :bows head: Any free time to put together a tutorial? I, as I am sure many others, will be eternally grateful. Question, how is software performance? Have you tested any "heavy software" software to see if they hang or lag? I remember reading a few threads on another forum with a successful dual core xeon build but within graphics and video editing software, performance was horrible. Let me know.

He'll hopefully post a guide when he finds time. In the meantime, you can use the msr199.kext workaround @catswift posted and get close to optimal performance it seems.

I'm consistently getting the same scores with Geekbench 3 and 4 that I posted before. For GPU performance see this post (a pretty hefty workaround for RX 480 in 10.12.1 beta's). I haven't tried any GPU intensive apps for a long stretch of time yet, unless you count Logic Pro X. There was a lot of speculation that it would or could utilize GPU's when the Mac Pro 2013 came out, but I'm not sure to what extent it does, if any. LPX performance has been superb though. I'm going to look into benchmarking that.
 
i have one 20core 2698V4 2.0ghz Engineering sample running on a Supermicro X10 SRAF since 2 weeks and testing... i have pretty impressive geek bench scores and all software runs flawlessly, no hangs, no lags etc...

However... when it comes to foto and Video editing, thats what i do, mainly in lightroom, and davinci resolve... those software cannot utilize the 40 threads available... in fact it looks like that most of this software can only keep 10-12 threads busy, if most... with than in mind the single core score actually defines the speeds you can achieve, not the multicore...

in my case with the 20core, i have a single core score comparable to my 4790K i had before so i see more or less the same speed s in this softwares... However i figured out that there is converter software than can scale to a unlimited amount of cores, in my case im using iFFMPEG to convert my 4K DCI h264 files to ProRes... this goes blazing fast, with about a 10 minutes file transcoded in 4 minutes (3x speed)... now that the files are in non CPU intensive formats, my single core speed is fast enough to have a blazing experience... To render timelines out in davinco, i achieved more or less 35-40 fps with 25fps 4K files... so more or less 1,5x... (on my 4790k i had 10-15 fps, but also there i used a 770GTX instead of a 980Ti im using now, so i blame the graphics card foremost not the cpu for the faster speed) i hope this will speed up once CUDA will be working on 10.12...


I have ordered another 20core 2698V4 2.0ghz Engineering sample, and a ASUS Z10PE-D8 motherboard... this should arrive mid next week to me so lets see how far can i push... once everything is running smooth and tested i will try to post some tutorial or similar...

Attached find some screenshots, the geek bench and cinebench was done under 10.11.6 but sierra has similar results...

A very interesting source for Multiprocessor Media software utilization is here:
https://www.pugetsystems.com/labs/a...o-CC-2015-Multi-Core-Performance-Update1-806/

This basically recommends you a lower core higher ghz CPU, wich all those xeon ES aren't...
 

Attachments

  • Bildschirmfoto 2016-09-30 um 18.05.13.png
    Bildschirmfoto 2016-09-30 um 18.05.13.png
    322.9 KB · Views: 229
  • Bildschirmfoto 2016-09-30 um 18.05.21.png
    Bildschirmfoto 2016-09-30 um 18.05.21.png
    173.7 KB · Views: 257
  • Bildschirmfoto 2016-09-28 um 04.22.17.jpg
    Bildschirmfoto 2016-09-28 um 04.22.17.jpg
    774.2 KB · Views: 236
I have had no trouble running everything I need, but I don't use the types of apps you've mentioned. Compilers (Xcode, mono, etc), Unity, Photoshop all working flawlessly. Most of the issues remaining are because of Sierra itself and not the hack (old webcam not working, old vpn no longer working). I compiled the latest Swift open source toolchain in 19 minutes (vs 3.5 hours it took a colleague on his mac pro).
 

Attachments

  • Screen Shot 2016-10-03 at 8.30.50 AM.png
    Screen Shot 2016-10-03 at 8.30.50 AM.png
    138.6 KB · Views: 193
No USB 3.0?
@volition Nice! I had tried the AppleALC kext but that was probably a day or two before @metacollin patch, will have to revisit.


For the BIOS settings for the USB audio fix, under Chipset Config -> South Bridge:

Legacy USB Support - Enabled
XCHI Hand-Off - Enabled
ECHI Hand-Off - Disabled
Port 60/64 - Disabled
USB 3.0 Support - Disabled:!:
EHCI1 - Enabled
EHCI2 - Enabled
Azalia - Disabled
 
No USB 3.0?

That's correct; for me all of my "usb 3" ports currently function as usb 2 ports. If you look further back in the posts though you'll note that this is my first hackintosh and that I have not tried to patch the ddst for this motherboard. I'm confident someone more experienced than I could get full usb 3 working ( **cough** @metacollin guide **cough** ).
 
Awesome information Panamamax! gold mine, especially that iffmpeg bit.

i have one 20core 2698V4 2.0ghz Engineering sample running on a Supermicro X10 SRAF since 2 weeks and testing... i have pretty impressive geek bench scores and all software runs flawlessly, no hangs, no lags etc...

However... when it comes to foto and Video editing, thats what i do, mainly in lightroom, and davinci resolve... those software cannot utilize the 40 threads available... in fact it looks like that most of this software can only keep 10-12 threads busy, if most... with than in mind the single core score actually defines the speeds you can achieve, not the multicore...

in my case with the 20core, i have a single core score comparable to my 4790K i had before so i see more or less the same speed s in this softwares... However i figured out that there is converter software than can scale to a unlimited amount of cores, in my case im using iFFMPEG to convert my 4K DCI h264 files to ProRes... this goes blazing fast, with about a 10 minutes file transcoded in 4 minutes (3x speed)... now that the files are in non CPU intensive formats, my single core speed is fast enough to have a blazing experience... To render timelines out in davinco, i achieved more or less 35-40 fps with 25fps 4K files... so more or less 1,5x... (on my 4790k i had 10-15 fps, but also there i used a 770GTX instead of a 980Ti im using now, so i blame the graphics card foremost not the cpu for the faster speed) i hope this will speed up once CUDA will be working on 10.12...


I have ordered another 20core 2698V4 2.0ghz Engineering sample, and a ASUS Z10PE-D8 motherboard... this should arrive mid next week to me so lets see how far can i push... once everything is running smooth and tested i will try to post some tutorial or similar...

Attached find some screenshots, the geek bench and cinebench was done under 10.11.6 but sierra has similar results...

A very interesting source for Multiprocessor Media software utilization is here:
https://www.pugetsystems.com/labs/a...o-CC-2015-Multi-Core-Performance-Update1-806/

This basically recommends you a lower core higher ghz CPU, wich all those xeon ES aren't...
 
@metacollin could you possibly provide the efi folder/config.plist for the X10Dai if its not too much to ask. Tried following your methods on my X10Dai and stuck at PCIConfig line. Also did u managed to update to Sierra on top of it? or still with 10.11.6?

thanks
 
@metacollin could you possibly provide the efi folder/config.plist for the X10Dai if its not too much to ask. Tried following your methods on my X10Dai and stuck at PCIConfig line. Also did u managed to update to Sierra on top of it? or still with 10.11.6?

thanks


I found this dude on Geekbench results browser who has a 68,000K hacktinosh. I then stumbled on a forum where he posts his config files but I am not a member of the forum to download. Perhaps we can figure out how to join and get the file see this here: (use google translate) http://bbs.pcbeta.com/viewthread-1712877-1-1.html
 
Status
Not open for further replies.
Back
Top