Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Status
The first post of this thread is a WikiPost and can be edited by anyone with the appropiate permissions. Your edits will be public.

beerstine

macrumors member
Jul 18, 2002
42
5
Di
Yeah, I definitely regret trying to do anything with the internal slot on the MP6,1. The most I ever got out of it was 1216.9MB/s / 1300.9 MB/s blackmagic benchmark with a Sabrent Rocket 4.

Meanwhile my thunderbolt array is pulling 2314.8MB/s / 2163.9 MB/s. Anandtech's system charts do seem indicate the internal slot has much less bandwidth than the PCIe switch in front of the Thunderbolt buses. I suppose I could toss the internal slot into the array next. Or try putting an array on the internal slot via an Angelshark.

Other regret is going with NVME cards and blades in my OWC Helios enclosures. It's easy to get a dual blade SATA card for cheap, and cheap SATA blades. I could have just had twice as many SATA blades instead of half as many NVME blades, and probably been far more compatible. E.g.:

Did you ever get an adapter to work in the 6,1? I'm not looking to RAID the SSD, just get a bigger and faster boot drive.
 

Tommi14

macrumors member
Jan 11, 2021
31
3
Hello,

Just to make sure: will these work in Mac Pro 4,1->5,1 (BootROM 144.0.0.0.0) running Mojave?


or


with


I suppose there is no big difference between SN750 and 970 EVO?

Thanks already,
Tommi
 
Last edited:

joevt

macrumors 604
Jun 21, 2012
6,967
4,262
Yeah, I definitely regret trying to do anything with the internal slot on the MP6,1. The most I ever got out of it was 1216.9MB/s / 1300.9 MB/s blackmagic benchmark with a Sabrent Rocket 4.
That seems reasonable. It's only a 2 GB/s link so I would expect something like 1500 MB/s. AmorphousDiskMark sequential test may get you closer to the actual limit.

Meanwhile my thunderbolt array is pulling 2314.8MB/s / 2163.9 MB/s. Anandtech's system charts do seem indicate the internal slot has much less bandwidth than the PCIe switch in front of the Thunderbolt buses. I suppose I could toss the internal slot into the array next. Or try putting an array on the internal slot via an Angelshark.

Other regret is going with NVME cards and blades in my OWC Helios enclosures. It's easy to get a dual blade SATA card for cheap, and cheap SATA blades. I could have just had twice as many SATA blades instead of half as many NVME blades, and probably been far more compatible. E.g.:
The Anandtech diagram is missing the link rate between the Thunderbolt controllers and the PEX8723 which are 2 GB/s each. And each TB2 controller should have two 2.5 GB/s downstream links (because each TB2 controller has two TB ports)
A TB2 controller can probably transmit 1.5 GB/s (both TB ports total) the other 3.5 GB/s can only be used up by DisplayPort (A 4K display is 2 GB/s). The DisplayPort data is input to the TB2 controllers so it doesn't take away from the upstream bandwidth.

So there should be a way to get a RAID 0 of ≈4 GB/s using one port from each TB2 controller. Connecting two NVMe SSDs to the same controller will slow both drives (if there total possible bandwidth exceeds the bandwidth of the controller) - and a RAID 0 is only as fast as its slowest drive (multiplied by the number of drives).
 

Generic_bug

macrumors newbie
Oct 9, 2018
9
2
Highpoint made SSD7103 for PCs, installed in a Mac Pro will work exactly like SSD7101A-1. SSD7101A-1 is bootable with a Mac Pro. It's the Mac Pro BootROM that makes it bootable, since 140.0.0.0.0, not the card firmware/kext/etc.

Last macOS release that can boot officially from an array is High Sierra. Apple still do not support bootable arrays with Big Sur.
Anyone knows WHY this absurd limitation happens ? It is so infuriating that Apple stopped supporting Booting from a RAID Array ... It limits you way too much... I suspect back in 2018 the reason was the switch to APFS .. but PLEASE ... This CANNOT BE STILL the reason... Any concrete answer ? Thanks a lot !
 

tsialex

Contributor
Jun 13, 2016
13,455
13,601
Anyone knows WHY this absurd limitation happens ? It is so infuriating that Apple stopped supporting Booting from a RAID Array ... It limits you way too much... I suspect back in 2018 the reason was the switch to APFS .. but PLEASE ... This CANNOT BE STILL the reason... Any concrete answer ? Thanks a lot !
Still the same answer, Apple did not implemented APFS bootable RAID support, even on 11.3b1.
 
  • Like
Reactions: Generic_bug

digiwolf

macrumors newbie
Feb 10, 2021
2
0
Hey All – Newbie here! I'm a bit confused when it comes to upgrading my MacPro 6,1 (Late 2013). I saw in this forum that Sabrent SSDs are great, however can I push to 2TB, instead of 1TB based on Apple's specs? Cheers!
 

Lance Nanek

macrumors member
Jun 23, 2020
31
5
Hey All – Newbie here! I'm a bit confused when it comes to upgrading my MacPro 6,1 (Late 2013). I saw in this forum that Sabrent SSDs are great, however can I push to 2TB, instead of 1TB based on Apple's specs? Cheers!
There's lots of 2TB blades listed as compatible on page 1. Capacity isn't super meaningful here. The controller chip used is far more important re if compatible. If an Phison E16 controller 1TB Sabrent works, a 2TB one will too.
 

digiwolf

macrumors newbie
Feb 10, 2021
2
0
There's lots of 2TB blades listed as compatible on page 1. Capacity isn't super meaningful here. The controller chip used is far more important re if compatible. If an Phison E16 controller 1TB Sabrent works, a 2TB one will too.

Thanks, Lance. I saw your post that a Sabrent 1TB is compatible, but I needed more. Will also buy the adaptor [Sintech NGFF M.2 nVME SSD Adapter Card] you recommended from Amazon. Thanks for your help!
 

tripmusic

macrumors 6502
Oct 26, 2012
455
86
Switzerland
I saw a review for the XPG SX8200 Pro and someone said they changed the controller late last year and it's 10-25% slower. Any truth to that?
 

Lance Nanek

macrumors member
Jun 23, 2020
31
5
My 256GB SX8200 Pro is very slow compared to the other blades I tested in Blackmagic. It's only 415 MB/s write, 1,086 MB/S read on Blackmagic set to 5GB in the expensive Glortrends enclosure. Some reviews seem to state the larger capacity versions didn't suffer from the change as much:
We also obtained sequential performance benchmarks with a 512GB model from a Redditor with the new SM2262G variant that comes with Samsung’s 64-Layer TLC flash. Again, our results are on the left and the Redditor’s results are on the right. At this capacity, the gaps are smaller

Meanwhile my 1TB Rocket 4 Plus in the JEYI enclosure does 1,217.8 MB/s write and 1287.7 MB/s read. So even though this is a Thunderbolt 2 Mac Pro 6,1; it is capable of more than the SX8200 Pro is giving it.

Meanwhile my Samsung PM981 256GB sticks are doing 1,516 MB/s write, 3,488 MB/s read each on CrystalDiskMark on my Windows 10 machine. So it's not impossible to find performant 256GB blades. Different benchmark apps, though. One may be exceeding any SLC cache these blades have, if any, whereas the other isn't.

Personally, I grabbed up a bunch of cheap used, 1TB, non-Q, Sabrent Rocket 3's for my array since no one wants them with the Rocket 4's out. Then I bought a bunch of cheap used Sabrent NVME Thunderbolt 3 enclosures that no one wants because they only do PCIe 3.0 x2 instead of x4 like almost every other enclosure on the market.

Haven't tried them on my Mac Pro 6,1 yet; but they are doing 992.5 MB/s write and 1,444.9 MB/s read on a Macbook Pro 2019. Hoping only having half the PCIe 3.0 lanes doesn't matter once I plugin to my Mac Pro 6,1 since it can only do PCIe 2.0 x 4 anyway. So no point paying for faster enclosures. I have no use for the capacity given the number of drives in my RAID 0 stripe array, but I started buying higher capacity blades just because they have larger SLC caches.
 

tsialex

Contributor
Jun 13, 2016
13,455
13,601
Hoping only having half the PCIe 3.0 lanes doesn't matter once I plugin to my Mac Pro 6,1 since it can only do PCIe 2.0 x 4 anyway. So no point paying for faster enclosures. I have no use for the capacity given the number of drives in my RAID 0 stripe array, but I started buying higher capacity blades just because they have larger SLC caches.
PCIe v2.0 x2 = 725~750MB/s
 

Lance Nanek

macrumors member
Jun 23, 2020
31
5
PCIe v2.0 x2 = 725~750MB/s
Does that matter? I was going to plug a PCIe v3.0 x2 enclosure into a Thunderbolt 3 dock to power it, then plug the dock into an Apple Thunderbolt 2<->3 adapter, then into the PCIe 2.0 x4 capable Mac Pro 6,1. Nothing in that should be running PCIe 2.0 x2, right?
 

tsialex

Contributor
Jun 13, 2016
13,455
13,601
Does that matter? I was going to plug a PCIe v3.0 x2 enclosure into a Thunderbolt 3 dock to power it, then plug the dock into an Apple Thunderbolt 2<->3 adapter, then into the PCIe 2.0 x4 capable Mac Pro 6,1. Nothing in that should be running PCIe 2.0 x2, right?
With just two PCIe lines, you will be limited to 725~750MB/s connected to a MP6,1.
 

Lance Nanek

macrumors member
Jun 23, 2020
31
5
With just two PCIe lines, you will be limited to 725~750MB/s connected to a MP6,1.
What makes you think it would just use 2?

Benchmarked just now at 1178.6 MB/s write and 1,276.6 MB/s read:
Mac Pro 6,1 -> Apple Thunderbolt 2 to 3 adapter -> Kensington SD5200T dock -> Sabrent EC-T3NS enclosure -> Sabrent Rocket 3 1TB NVME

NVME Diagnostics:
Code:
Sabrent:

  Capacity:    1.02 TB (1,024,209,543,168 bytes)
  TRIM Support:    Yes
  Model:    Sabrent
  Revision:    RKT303.3
  Serial Number:    C1D30702120901088389
  Link Width:    x2
  Link Speed:    8.0 GT/s
  Detachable Drive:    No
  BSD Name:    disk8
  Partition Map Type:    GPT (GUID Partition Table)
  Removable Media:    No
  S.M.A.R.T. status:    Verified
  Volumes:
EFI:
  Capacity:    209.7 MB (209,715,200 bytes)
  File System:    MS-DOS FAT32
  BSD Name:    disk8s1
  Content:    EFI
  Volume UUID:    0E239BC6-F960-3107-89CF-1C97F78BB46B
SABRENT_RKT3:
  Capacity:    1.02 TB (1,023,865,569,280 bytes)
  Free:    1.02 TB (1,023,201,193,984 bytes)
  Writable:    Yes
  File System:    Journaled HFS+
  BSD Name:    disk8s2
  Mount Point:    /Volumes/SABRENT_RKT3
  Content:    Apple_HFS
  Volume UUID:    128C7647-0F32-30E9-99D5-94CF475550B2

Thunderbolt bus:
Code:
Thunderbolt Bus 0:

  Vendor Name:    Apple Inc.
  Device Name:    Mac Pro
  UID:    0x0001000E01AD9E30
  Route String:    0
  Firmware Version:    19.2
  Domain UUID:    40645F84-4600-1250-88CB-7B3D297E2DCF
  Port:
  Status:    Device connected
  Link Status:    0x2
  Speed:    Up to 20 Gb/s x1
  Current Link Width:    0x2
  Receptacle:    6
  Cable Firmware Version:    1.0.16
  Cable Serial Number:    C4M032504FLF798AT
  Link Controller Firmware Version:    0.13.0
  Port:
  Status:    Device connected
  Link Status:    0x2
  Speed:    Up to 20 Gb/s x1
  Current Link Width:    0x2
  Receptacle:    5
  Cable Firmware Version:    1.0.16
  Cable Serial Number:    C4M04650228F798AZ
  Link Controller Firmware Version:    0.13.0

Thunderbolt3 SSD Dock:

  Vendor Name:    Guangzhou JEYI Information Technology CO., LTD
  Device Name:    Thunderbolt3 SSD Dock
  Vendor ID:    0x265
  Device ID:    0x1
  Device Revision:    0x1
  UID:    0x02654DFC28085300
  Route String:    1
  Firmware Version:    59.1
  Port (Upstream):
  Status:    Device connected
  Link Status:    0x2
  Speed:    Up to 20 Gb/s x1
  Current Link Width:    0x2
  Cable Firmware Version:    1.0.16
  Cable Serial Number:    C4M032504FLF798AT
  Link Controller Firmware Version:    1.42.0
  Port:
  Status:    No device connected
  Link Status:    0x7
  Speed:    Up to 40 Gb/s x1
  Current Link Width:    0x1
  Link Controller Firmware Version:    1.42.0

SD5200T:

  Vendor Name:    Kensington Computer Products Group
  Device Name:    SD5200T
  Vendor ID:    0x1B
  Device ID:    0x2
  Device Revision:    0x2
  UID:    0x001BCB4C9C324200
  Route String:    3
  Firmware Version:    23.1
  Port (Upstream):
  Status:    Device connected
  Link Status:    0x2
  Speed:    Up to 20 Gb/s x1
  Current Link Width:    0x2
  Cable Firmware Version:    1.0.16
  Cable Serial Number:    C4M04650228F798AZ
  Link Controller Firmware Version:    0.26.0
  Port:
  Status:    Device connected
  Link Status:    0x2
  Speed:    Up to 40 Gb/s x1
  Current Link Width:    0x2
  Link Controller Firmware Version:    0.26.0

Sabrent Enclosure:

  Vendor Name:    Sabrent
  Device Name:    Sabrent Enclosure
  Vendor ID:    0x1DF
  Device ID:    0x107
  Device Revision:    0x1
  UID:    0x01DF6ADD14BB2A00
  Route String:    103
  Firmware Version:    20.1
  Port (Upstream):
  Status:    Device connected
  Link Status:    0x2
  Speed:    Up to 40 Gb/s x1
  Current Link Width:    0x2
  Link Controller Firmware Version:    0.31.0

Not sure which of these NVME controllers it is, but think it is one of the x2 one:
Code:
pci1cc1,8201:

  Type:    NVM Express Controller
  Driver Installed:    Yes
  Tunnel Compatible:    Yes
  Pause Compatible:    Yes
  MSI:    Yes
  Bus:    PCI
  Slot:    Thunderbolt@83,0,0
  Vendor ID:    0x1cc1
  Device ID:    0x8201
  Subsystem Vendor ID:    0x1cc1
  Subsystem ID:    0x8201
  Revision ID:    0x0003
  Serial Number:    2K472LQ46EJJ
  Link Width:    x4
  Link Speed:    8.0 GT/s

pci1987,5012:

  Type:    NVM Express Controller
  Driver Installed:    Yes
  Tunnel Compatible:    Yes
  Pause Compatible:    Yes
  MSI:    Yes
  Bus:    PCI
  Slot:    Thunderbolt@220,0,0
  Vendor ID:    0x1987
  Device ID:    0x5012
  Subsystem Vendor ID:    0x1987
  Subsystem ID:    0x5012
  Revision ID:    0x0001
  Serial Number:    C1D30702120901088389
  Link Width:    x2
  Link Speed:    8.0 GT/s

pci1987,5012:

  Type:    NVM Express Controller
  Driver Installed:    Yes
  Tunnel Compatible:    Yes
  Pause Compatible:    Yes
  MSI:    Yes
  Bus:    PCI
  Slot:    Thunderbolt@220,0,0
  Vendor ID:    0x1987
  Device ID:    0x5012
  Subsystem Vendor ID:    0x1987
  Subsystem ID:    0x5012
  Revision ID:    0x0001
  Serial Number:    C1D30702120901088389
  Link Width:    x2
  Link Speed:    8.0 GT/s
 

Attachments

  • benchmark-SD5200T_dock-Sabrent_EC-T3NS_enclosure-Sabrent_RKT3_NVME.png
    benchmark-SD5200T_dock-Sabrent_EC-T3NS_enclosure-Sabrent_RKT3_NVME.png
    221.2 KB · Views: 122

Lance Nanek

macrumors member
Jun 23, 2020
31
5
Maybe there was a double in that last block, these are the 3 NVME controllers:
Code:
pci1cc1,8201:

  Type:    NVM Express Controller
  Driver Installed:    Yes
  Tunnel Compatible:    Yes
  Pause Compatible:    Yes
  MSI:    Yes
  Bus:    PCI
  Slot:    Thunderbolt@83,0,0
  Vendor ID:    0x1cc1
  Device ID:    0x8201
  Subsystem Vendor ID:    0x1cc1
  Subsystem ID:    0x8201
  Revision ID:    0x0003
  Serial Number:    2K472LQ46EJJ
  Link Width:    x4
  Link Speed:    8.0 GT/s

pci1987,5012:

  Type:    NVM Express Controller
  Driver Installed:    Yes
  Tunnel Compatible:    Yes
  Pause Compatible:    Yes
  MSI:    Yes
  Bus:    PCI
  Slot:    Thunderbolt@220,0,0
  Vendor ID:    0x1987
  Device ID:    0x5012
  Subsystem Vendor ID:    0x1987
  Subsystem ID:    0x5012
  Revision ID:    0x0001
  Serial Number:    C1D30702120901088389
  Link Width:    x2
  Link Speed:    8.0 GT/s

pci1987,5018:

  Type:    NVM Express Controller
  Driver Installed:    Yes
  Tunnel Compatible:    Yes
  Pause Compatible:    Yes
  MSI:    Yes
  Bus:    PCI
  Slot:    Thunderbolt@237,0,0
  Vendor ID:    0x1987
  Device ID:    0x5018
  Subsystem Vendor ID:    0x1987
  Subsystem ID:    0x5018
  Revision ID:    0x0001
  Serial Number:    6D12070B14A292137352
  Link Width:    x4
  Link Speed:    8.0 GT/s
 

tsialex

Contributor
Jun 13, 2016
13,455
13,601
Maybe there was a double in that last block, these are the 3 NVME controllers:
Code:
pci1cc1,8201:

  Type:    NVM Express Controller
  Driver Installed:    Yes
  Tunnel Compatible:    Yes
  Pause Compatible:    Yes
  MSI:    Yes
  Bus:    PCI
  Slot:    Thunderbolt@83,0,0
  Vendor ID:    0x1cc1
  Device ID:    0x8201
  Subsystem Vendor ID:    0x1cc1
  Subsystem ID:    0x8201
  Revision ID:    0x0003
  Serial Number:    2K472LQ46EJJ
  Link Width:    x4
  Link Speed:    8.0 GT/s

pci1987,5012:

  Type:    NVM Express Controller
  Driver Installed:    Yes
  Tunnel Compatible:    Yes
  Pause Compatible:    Yes
  MSI:    Yes
  Bus:    PCI
  Slot:    Thunderbolt@220,0,0
  Vendor ID:    0x1987
  Device ID:    0x5012
  Subsystem Vendor ID:    0x1987
  Subsystem ID:    0x5012
  Revision ID:    0x0001
  Serial Number:    C1D30702120901088389
  Link Width:    x2
  Link Speed:    8.0 GT/s

pci1987,5018:

  Type:    NVM Express Controller
  Driver Installed:    Yes
  Tunnel Compatible:    Yes
  Pause Compatible:    Yes
  MSI:    Yes
  Bus:    PCI
  Slot:    Thunderbolt@237,0,0
  Vendor ID:    0x1987
  Device ID:    0x5018
  Subsystem Vendor ID:    0x1987
  Subsystem ID:    0x5018
  Revision ID:    0x0001
  Serial Number:    6D12070B14A292137352
  Link Width:    x4
  Link Speed:    8.0 GT/s
If you connect without the dock, how much you get? Something on your chain is acting like a switch. Your dock has a TB switch?
 

Lance Nanek

macrumors member
Jun 23, 2020
31
5
If you connect without the dock, how much you get? Something on your chain is acting like a switch. Your dock has a TB switch?
I don't think it would work without the dock. The enclosure is bus powered, but the Apple Thunderbolt 2 to 3 adapter doesn't provide anyway.

The three TB2 bus controllers are behind a PEX 8723. So there is a switch in there somewhere.
 

Attachments

  • MPsystemarch_north.png
    MPsystemarch_north.png
    53.4 KB · Views: 125

YosemiteSam

macrumors member
Aug 26, 2010
48
4
Southern California
Would appreciate some advice before I make some NVMe purchases.

I'm looking primarily at the Highpoint SSD 7104 (the fan-less 7101-A1). The seemingly less-likely option is the Sonnet M.2 4x4, but I recall others indicating that that this card doesn't physically fit with one or more of the PCIe cards I also happen to have (too big for the space). Feedback/advice for my situation is much appreciated.

Background and Use Case
I'm a composer, mostly working with midi/samples/audio. I use a 2010 Mac Pro 5,1, currently running 10.3 HS (planning to move to 10.4 Mojave soon). Don't plan to move beyond Mojave unless OpenCore proves to be *SUPER* stable, compromise-free, and incredibly easy to manage in a pro DAW setup. I will need to replace my machine in the coming years (already been in a "holding pattern" for 2+ years), but am currently looking to extend its usable life for at least another year (or two or three) while many things get sorted out with Apple's pro workstations (namely, price/value/lifespan and what happens at the "pro" level as things switch to Apple silicon). I currently need to expand my samples storage, but want to make a choice that will work well with my current setup AND allow for a seamless transition to my future Mac setup (ideally physically swap from one machine to another, so what I invest in the 5,1 is also an investment into my future machine). Top priorities are reliability, performance (most likely non-RAIDed), quietness, and compatibility/transferability between my 5,1 and a future 7,1 (or 8,1 etc.)

Current PCIe setup
Slot 4: MOTU PCIe-424 audio card (hosts my audio interfaces)
Slot 3: Sonnet Tempo SSD Pro Plus (2 x Samsung 850 EVO SSDs), also connecting via eSATA to external Firmtek Seritek/5PM enclosure with spinning HDDs.
Slot 2: empty
Slot 1: Sapphire Pulse RX 580 video card
All hard drive slots are used- some SSD, some HDD.

Ideally, I'd get a card/storage that works perfectly in that 2nd (x16) slot. I plan to use Samsung blades (likely 970 EVO Plus). Non-RAID at this point, but would like to have the option for RAID in the future.

Will the Highpoint SSD 7104 work for my purposes, and will it physically fit and play nicely between the RX 580 and the Sonnet SSD card?
 

joevt

macrumors 604
Jun 21, 2012
6,967
4,262
If you connect without the dock, how much you get? Something on your chain is acting like a switch. Your dock has a TB switch?
The devices are like this:
Code:
Thunderbolt Bus 0:
	downstream port 1:
		Thunderbolt3 SSD Dock (Guangzhou JEYI Information Technology CO., LTD)
		Speed:    Up to 20 Gb/s x1
		Current Link Width:    0x2
	downstream port 2:
		SD5200T (Kensington Computer Products Group)
		Speed:    Up to 20 Gb/s x1
		Current Link Width:    0x2
			downstream port:
				Speed:    Up to 40 Gb/s x1
				Current Link Width:    0x2
				Sabrent Enclosure (Sabrent)
All of them (except the Bus) are Thunderbolt 3 devices, so they all use PCIe 3.0 link rate. The Thunderbolt controllers in each Thunderbolt Bus and device acts as a PCIe switch. The upstream of the Thunderbolt controller (Thunderbolt Bus) in the MacPro6,1 is PCIe 2.0 x4 (a real PCIe connection). The downstreams are going to be PCIe 1.0 x4 for fake PCIe (Thunderbolt connection 20 Gb/s to the MacPro6,1 or 40 Gb/s between two Thunderbolt 3 devices) or PCIe 3.0 x4 or x2 for real PCIe (NVMe connection).

See the attachment at https://forums.macrumors.com/threads/testing-tb3-aic-with-mp-5-1.2143042/post-27612591 for examples.

If the Sabrent Enclosure could work without the Kensington dock, it would still use PCIe 3.0 link rate because of its Thunderbolt 3 controller.
 

joevt

macrumors 604
Jun 21, 2012
6,967
4,262
Will the Highpoint SSD 7104 work for my purposes, and will it physically fit and play nicely between the RX 580 and the Sonnet SSD card?
I think fanless will fit? I have an SSD7505 with fan and it doesn't fit (the fan sticks out a bit so it won't turn because it makes contact with the PCIe card in slot 3). https://forums.macrumors.com/threads/pcie-ssds-nvme-ahci.2146725/post-29483973

I guess they should all play nicely together. You might have problems if you try to boot Windows, unless you boot Windows using EFI which has its own problems.
 

mayuka

macrumors 6502a
Feb 15, 2009
610
66
In my 2019 Mac Pro I have two Wolftech Pulsecard adapters, each with a Samsung EVO 970 Pro 2 TB SSD. While one drive (the second one) mounts without any issues, the first one (which contains the Time Machine Backup) only mounts with a timeout of 2-5 minutes after each boot. I have the feeling that these timeouts are getting worse (longer). I can't detect anything in the log files and SMART values are all ok (tested with smartctl and Disk Utils). The second drive seems to have very similar values reported. I am still on 10.15.7.

Any ideas on these timeouts? Could it be a failing SSD and/or failing PCIe adapter?

Code:
Generic SSD Controller:

Samsung SSD 970 EVO Plus 2TB:

  Capacity:    2 TB (2.000.398.934.016 bytes)
  TRIM Support:    Yes
  Model:    Samsung SSD 970 EVO Plus 2TB
  Revision:    2B2QEXM7
  Serial Number:    -------------------
  Link Width:    x4
  Link Speed:    8.0 GT/s
  Detachable Drive:    No
  BSD Name:    disk1
  Partition Map Type:    GPT (GUID Partition Table)
  Removable Media:    No
  SMART status:    Verified
  Volumes:
EFI:
  Capacity:    209,7 MB (209.715.200 bytes)
  File System:    MS-DOS FAT32
  BSD Name:    disk1s1
  Content:    EFI
  Volume UUID:    0E239BC6-F960-3107-89CF--------------
disk1s2:
  Capacity:    2 TB (2.000.189.177.856 bytes)
  BSD Name:    disk1s2
  Content:    Apple_APFS


# smartctl -a disk1
smartctl 7.1 2019-12-30 r5022 [Darwin 19.6.0 x86_64] (local build)
Copyright (C) 2002-19, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF INFORMATION SECTION ===
Model Number:                       Samsung SSD 970 EVO Plus 2TB
Serial Number:                      ------------------
Firmware Version:                   2B2QEXM7
PCI Vendor/Subsystem ID:            0x144d
IEEE OUI Identifier:                0x002538
Total NVM Capacity:                 2,000,398,934,016 [2.00 TB]
Unallocated NVM Capacity:           0
Controller ID:                      4
Number of Namespaces:               1
Local Time is:                      Sun Feb 14 10:04:55 2021 CET
Firmware Updates (0x16):            3 Slots, no Reset required
Optional Admin Commands (0x0017):   Security Format Frmw_DL Self_Test
Optional NVM Commands (0x005f):     Comp Wr_Unc DS_Mngmt Wr_Zero Sav/Sel_Feat Timestmp
Maximum Data Transfer Size:         512 Pages
Warning  Comp. Temp. Threshold:     85 Celsius
Critical Comp. Temp. Threshold:     85 Celsius

Supported Power States
St Op     Max   Active     Idle   RL RT WL WT  Ent_Lat  Ex_Lat
0 +     7.50W       -        -    0  0  0  0        0       0
1 +     5.90W       -        -    1  1  1  1        0       0
2 +     3.60W       -        -    2  2  2  2        0       0
3 -   0.0700W       -        -    3  3  3  3      210    1200
4 -   0.0050W       -        -    4  4  4  4     2000    8000

=== START OF SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED

SMART/Health Information (NVMe Log 0x02)
Critical Warning:                   0x00
Temperature:                        34 Celsius
Available Spare:                    100%
Available Spare Threshold:          10%
Percentage Used:                    1%
Data Units Read:                    354,370,948 [181 TB]
Data Units Written:                 81,436,633 [41.6 TB]
Host Read Commands:                 3,701,804,022
Host Write Commands:                643,633,980
Controller Busy Time:               2,677
Power Cycles:                       215
Power On Hours:                     978
Unsafe Shutdowns:                   21
Media and Data Integrity Errors:    0
Error Information Log Entries:      0
Warning  Comp. Temperature Time:    0
Critical Comp. Temperature Time:    0
Temperature Sensor 1:               34 Celsius
Temperature Sensor 2:               35 Celsius

Read Error Information Log failed: NVMe admin command:0x02/page:0x01 is not supported
 
Last edited:

Nightowl23

macrumors newbie
Feb 14, 2021
2
0
At the moment, it's the only two confirmed working on a Mac Pro. Other cards with ASM2824 PCI-Express 3.0 x24 switch could work, but someone has to buy one and test it for the rest of us :p

This is a good candidate:

SYBA Dual M.2 NVMe Ports to PCIe 3.0 x16 Bifurcation Riser Controller - Support Non-BiFurcation Motherboard, SI-PEX40129
Tsialex,
Obviously I'm a newbie but love (and hate) working on my (2012) 5,1. I'm trying to boot Windows from a dedicated NVME blade. I've heard lots of No's but ran across a few Yes's. Can you suggest a guide that would allow me to do this even if I had to PRAM on every boot. The Mac OS would be on the internal SSD Sam 960 and the NVME would WD N750 in a PCIE slot.
Thanks in advance for any help!
 

tsialex

Contributor
Jun 13, 2016
13,455
13,601
Tsialex,
Obviously I'm a newbie but love (and hate) working on my (2012) 5,1. I'm trying to boot Windows from a dedicated NVME blade. I've heard lots of No's but ran across a few Yes's. Can you suggest a guide that would allow me to do this even if I had to PRAM on every boot. The Mac OS would be on the internal SSD Sam 960 and the NVME would WD N750 in a PCIE slot.
Thanks in advance for any help!
No one should install UEFI Windows before knowing the health of your Mac Pro BootROM and without OC SecureBoot protection (see the OpenCore thread on the stickies). You could be on the verge of a brick.

I'll send you a PM.
 
  • Like
Reactions: trifero

dmMacRumors

macrumors newbie
Jan 23, 2021
4
0
Sabrent Rocket 4 Plus runs fine on the internal slot in Mac Pro 6,1 with the Sintech-C adapter and Catalina OS. Showing x4 5.0 GT/s which is as fast as it gets on this machine (PCIe 2.0).

Returned the Sabrent Rocket 4 I was trying to get running in an OWC Mercury Helios Thunderbolt 2 enclosure. Never got it to run more than x4 2.5 GT/s there (PCIe 1.0). Ran fine at x4 8.0 GT/s in a Glortrends Thunderbolt 3 enclosure. So I imagine you would get that speed with PCIe 3.0.
Did you find the Sintech-C adaptor a hard fit to hold a double sided SSD? Sintech on Amazon seems to indicate it's a problem for the Sintech-D adaptor which I think is just the C adaptor plus tools. It's in the description plus in one of the photos. They say nothing on Sintech's site though. https://www.amazon.com/Sintech-Comp...&keywords=ST-NGFF2013-D&qid=1613161137&sr=8-1

I've yet to commit my order for that drive plus adaptor but getting close. Did you ever note power consumption when benchmarking? My guess is that if the power management of the SSD firmware is well tuned, you should come in lower than other SSD's when running PCIe3.0 or 2.0, as the chip is fabbed on 12nm FinFet and descriptions on Sabrents site indicates need a heatsink if on PCIe4.0 and they say nothing otherwise.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.