Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

NVMe with ST-NGFF2013-C; Vega Internal GPU; Mac Pro 2013 (6,1)

CodeJingle said:
I didn't know we were quoting from the documentation that is supposed to be restricted access. It might be a better idea to use subtlety when referencing official Apple service manuals or schematics to not get on Apple's naughty list. It doesn't seem like something an Apple employee would do.
Click to expand...

One does not have to be an Apple employee to get the service manual. You simply need to be with an AASP and have GSX access.
 
  • Like
Reactions: itdk92 and dabotsonline
rw3 said:
One does not have to be an Apple employee to get the service manual. You simply need to be with an AASP and have GSX access.
Click to expand...
You are missing my point. Is someone with GSX access free to discuss the content of a service manual with someone who does not have GSX access? If Apple allows someone with GSX access to freely quote and directly reference the entire contents of the service manual to people without GSX access, then what is the purpose of GSX access in the first place? It makes no sense then; the documents should be freely available.

Or maybe Apple does not allow someone with GSX access to freely discuss the content of service manuals with those who do not have access. In which case, they should be more subtle. Apple Authorized Service Provider is still controlled by Apple. Apple still has provisions and restrictions on content retrieved in a manner that required GSX access and what portion of that material is allowed to be disclosed to those without GSX access. At best someone who wanted to break protocol would do so in a private message rather than disclosure to an entire thread.
 
Last edited:
CodeJingle said:
You are missing my point. Is someone with GSX access free to discuss the content of a service manual with someone who does not have GSX access? If Apple allows someone with GSX access to freely quote and directly reference the entire contents of the service manual to people without GSX access, then what is the purpose of GSX access in the first place? It makes no sense then, the documents should be freely available. Or maybe Apple does not allow someone with GSX access to freely discuss the content of service manuals with those who do not have access. In which case, they should be more subtle.
Click to expand...

Says the guy who is attempting to produce GPUs with a pinout that hasn’t been able to be obtained for over 4 years....
 
  • Like
Reactions: dabotsonline
rw3 said:
Says the guy who is attempting to produce GPUs with a pinout that hasn’t been able to be obtained for over 4 years....
Click to expand...
I've settled on using continuity testing from the empty CPU socket. And since the pinout of the Xeon processor is public documentation I'm in the clear.

Look I just don't want anyone to get in trouble. If you have dreams of working at Apple, then be more subtle with sharing or directly quoting from documentation that requires special access. That is all I am saying.
 
Last edited:
CodeJingle said:
For me no only the size. I said earlier in the thread I had 1.5 TB internal storage in my 2010 MacBook Pro I've been waiting for four years to go back to a larger size, 2 TB is only marginally better. I already had 1 TB in my Mac Pro before the upgrade. $1,200 USD is a lot of money but I literally had no choice here.
Click to expand...
Specific applications or work needing much space? For myself the upgrade is not a necessity, I have a Black Magic TB2 drivedock loaded with SSDs at home. But for those moments when I take my Mac Pro with me, the 256GB internal drive is proving too small.
 
jclmavg said:
Specific applications or work needing much space? For myself the upgrade is not a necessity, I have a Black Magic TB2 drivedock loaded with SSDs at home. But for those moments when I take my Mac Pro with me, the 256GB internal drive is proving too small.
Click to expand...
I go back and forth between work and home on a daily basis and I would prefer not have a bunch of external drives as I'm likely to eventually forget them (I sometimes even forget my Mac Pro).
[doublepost=1510881276][/doublepost]
dabotsonline said:
Isn't Flex / Flextronics the contract manufacturer / Electronics Manufacturing Service (EMS) for the Mac Pro 2013, at their factory in Austin, Texas?
Click to expand...
My mistake I'll update my original post.
 
dabotsonline said:
Isn't Flex / Flextronics the contract manufacturer / Electronics Manufacturing Service (EMS) for the Mac Pro 2013, at their factory in Austin, Texas?
Click to expand...

I would consider this 1st party chain of custody. Flextronics simply assembles the Mac Pro from parts that are manufactured elsewhere. I've been in this factory and they do nothing more than put all the parts together.
[doublepost=1510884419][/doublepost]
CodeJingle said:
I've settled on using continuity testing from the empty CPU socket. And since the pinout of the Xeon processor is public documentation I'm in the clear.
Click to expand...

If you are going to do the testing from the empty CPU socket, you are still going to need the logic board. The logic board is a glorified interconnect board that has the chipset on it, so you will need it to understand how the GPU and/or I/O board cables work and are pinned.
 
RyanXM said:
If you are going to do the testing from the empty CPU socket, you are still going to need the logic board.
Click to expand...

I am doing continuity testing from the empty CPU socket to every other connector pin on the Mac Pro, which requires it to be practically fully assembled. CPU socket pin 1 is one end, graphics card A connector pin 1 is the other end, for example. The CPU is not in its socket, but the CPU board as a whole is still plugged into the logic board. Since my new method of testing is no longer destructive, I don't need to buy another Mac Pro.
 
Last edited:
RyanXM said:
If you are going to do the testing from the empty CPU socket, you are still going to need the logic board. The logic board is a glorified interconnect board that has the chipset on it, so you will need it to understand how the GPU and/or I/O board cables work and are pinned.
Click to expand...
I agree - luckily for the graphics cards there aren’t additional controllers in-between.

MPsystemarch_north.png
 
MarkJames68 said:
RyanXM said:
you are still going to need the logic board
Click to expand...
I agree
Click to expand...

I have no idea how I would do continuity testing without the logic board, or without the rest of the computer. I need most or all of the computer to do continuity testing from the CPU socket. I'm using my own machine for testing. I'll start this weekend, maybe even Friday night. I'm hoping to post a video of the process so that should clear up any confusion. I'm also hoping I don't accidentally break my computer in the process. It takes very little circuitry to get in the way of a continuity test so I'm still worried it doesn't work.
 
Last edited:
RyanXM said:
I would imagine the chipset is still in the middle of some of this....
Click to expand...

Even in the ideal case, the Xeon CPU socket has 2,011 connectors. I'll get to the chipset eventually. I have to start somewhere. Unless all 2,011 connections also go through the chipset.
...
After someone knawing on my ankles to do this, I bought a spare logic board to tear apart if it is helpful to map between the CPU and the chipset before going out to the graphics cards. It may be useful in the halfway mapping between CPU <=> chipset <=> graphics boards a+b. Although I am beyond broke at this point.

Screen Shot 2017-11-17 at 3.36.58 PM.png
 
Last edited:
The 2,011 connectors are not all labeled on a single page, which is the 'cheat sheet' I need to make this easier (or even possible). Otherwise Intel's official documentation uses a seemingly abstract concept of a 'land' for connector grouping (and there are 50 of these groups or 'lands').

Does anyone have a single-page document or image for the Xeon-E5-v2 with all 2,011 connectors labeled? Maybe the mechanical model step file or the bsdl file contains this.

https://www.intel.com/content/dam/www/public/us/en/documents/datasheets/xeon-e5-v2-datasheet-vol-1.pdf

https://www.intel.com/content/dam/w...nts/datasheets/xeon-e5-v2-datasheet-vol-2.pdf

71FC8F72-2F83-47E8-A280-F5D7F6636B1A.jpeg
 
Last edited:
Spending a couple days looking for a dumbed-down 2011 connector pinout and so far it literally doesn't exist. I've blown it up so I can fill it in myself. I'm just kind of flabbergasted with all the years of complaining about Mac Pro 2013 that I am the first one trying this. I'll post again after filling it out. I am probably waiting for the extra logic board to come in the mail before doing the continuity testing.

18dQ9qggZ0zBQ40-DwFWtaq7KONqtytB7Ei8kI1OEbA,SF2dvPUrGNZRm9s2OneNEXuH3q6EV2JWIvn27oX1rHs-1.jpg
 
Last edited:
Last edited:
As an Amazon Associate, MacRumors earns a commission from qualifying purchases made through links in this post.
I'll need to eventually figure out which pins to re-route to the Vega graphics card (or a similar graphics card compatible with PCIe 3.0 x 16). To make that process simpler, I found a regular Xeon E5 v2 board on the cheap from eBay. I can do a similar continuity test on that board with the graphics card plugged in (or just continuity test from the CPU slot or chipset to the empty PCIe slot pins) and it should help fill in the gaps compared to the continuity test on the Mac Pro.

XeonBoard.png

[doublepost=1511160560][/doublepost]
IBICO said:
Adapter on Amazon say Currently unavailable.
https://www.dustinhome.no/product/5010982125/960-pro
Click to expand...

I know it looks like sweet potatoes, ignore that part. This is the webpage of the 'PC PARTS 239'-branded adapter I bought. Maybe you can buy it directly from the manufacturer?

http://www.pcparts239.com/?pid=123383677
 
Last edited:
okay, I just received the latest adapter from SINTECH. an otherwise perfectly working SAMSUNG 960 EVO again was NOT recognized by the Mac Pro. that's it for me. luckily, I do own a bunch of spare APPLE/SAMSUNG SSUBXs. will not investigate further...
 
Last edited:
mikeboss said:
okay, I just received the latest adapter from SINTECH. an otherwise perfectly working SAMSUNG 960 EVO again was NOT recognized by the Mac Pro. that's it for me. luckily, I do own a bunch of spare APPLE/SAMSUNG SSUBXs. will not investigate further...
Click to expand...
Ok, I find this somewhat alarming, I am waiting for that adapter too. You are on High Sierra, right?
 
mikeboss said:
okay, I just received the latest adapter from SINTECH. an otherwise perfectly working SAMSUNG 960 EVO again was NOT recognized by the Mac Pro. that's it for me. luckily, I do own a bunch of spare APPLE/SAMSUNG SSUBXs. will not investigate further...
Click to expand...
On this thread collectively we only know the PC Parts 239 adapter as working, first claimed by me and verified by MarkJames68. You were fine making the assumption earlier to buy a different adapter based on a known brand and international availability, before anyone verified the PC Parts 239 connector. But it seems kind of loopy to give up now when we know a specific adapter is working. You didn't ask the thread if someone from America would help you out. If I can acquire another one I'll be glad to ship you one of these adapters. I'm not going to do this for everyone but not a big deal to do it for one or two people.
 
Last edited:
  • Like
Reactions: ActionableMango
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back