Are you using a 10 Gbps port of the TS3+ Dock (there's one USB-C and one Thunderbolt)? Or are you using one of the 5 Gbps ports?
https://www.caldigit.com/ts3-plus-interface-bandwidth-allocation-and-diagram/
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
USB on M1 Macs isn't actually 10Gb/s? (Also definitely not USB4)
- Thread starter Anonymous Freak
- Start date
-
- Tags
- apple silicon mac mini usb 3.2
- Sort by reaction score
https://www.caldigit.com/ts3-plus-interface-bandwidth-allocation-and-diagram/
Yes, I test at this ports.
Tomorrow make a video showing speeds, cable, enclosure and porst
This is nothing to do with Apple but the T5/7 Samsung drives. They are not optimised for sustained throughout rather small reads/writes. There are a lot of articles about this problem.
My test had nothing to do with the actual throughput the drives are capable of, and everything to do with the USB link speed they negotiate.
As @Anonymous Freak said, this is about the USB connection link speed that the M1 negotiates with the drives (seen through the MacOS system report, under the USB section) .
I have both the T5 and T7 drives and can confirm that the T5 only connects at 5Gbps to the M1, but at 10Gbps on an Intel Mac (MBP16). The T7 connects at 10Gbps on both M1 and Intel Macs.
If I connect the T5 via a TB3 dock (CalDigit Plus), it connects at 10Gbps.
However, throughput speeds (with BlackMagic Disk Speed Test) show that both T5 and T7 are faster when connected via the TB3 dock.
In the case of the T7 it is about 55-60MBps faster via the dock (but still 100-200MBps slower than when connected to an Intel Mac)
The T5 difference is much greater because it only connects at 5Gbps, and gets 350/385MBps read/write, vs 480/530MBps read/write via the dock.
Interestingly, there is only minimal difference when connecting to the T5 to the MBP16 (which connects at 10Gbps) - about 10MBps faster in write. I expect the T5 is hitting the SATA3 speed limit, as I get an almost identical result with another SATA3 SSD in a 10Gbps enclosure.
So the M1 Macs have two problems:
1) Many disks that support USB 3.1 gen 2 at 10Gbps on an Intel Mac only connect at 5Gbps (I have tested about 6 personally)
2) Even when connected at 10Gbps, the M1 transfer speeds are significantly lower than Intel Macs.
Yeah, T5 can’t do much faster than ~550 MB/s due to SATA limitation. It unfortunate M1’s USB-C/TB speed is subpar.
The same is true for 5Gbps connection also (whether the device is 5Gbps only or 10Gbps connected at 5Gbps). This means the problem is not the bandwidth of the port - but the USB controller's handling of the USB protocol itself.
In contrast, the problem with a USB controller like the ASM1142 is that it is limited to 8Gbps which means it can do 5 Gbps just fine - such connections are not be limited, but a 10 Gbps connection will be limited.
Has anyone already tested this with a 2021 M1 iMac? According to the iFixit teardown some of the components / controller chips seem to have slightly different model numbers now compared to the „first wave“ of M1 devices from late 2020.
I've bought an M1 iMac and my UGreen USB 3.1 2 gen enclosure with a Plextor PX-256M5S SSD shows 10 Gb/s when connected to the USB 4 ports but only 5b/s when connected to the Thunderbolt ports.
However the read/write speed is rather disappointing :/
However the read/write speed is rather disappointing :/
Wow! That’s weird. Seems almost like the M1 MacBooks where the TB ports are not being recognized as 10 Mbps in some cases.
Edit: and I think you mean when connected to the USB 3 ports. The USB4 ports are the Thunderbolt ports.
We've got a confusing mix of ports on Macs, haven't we!
At least the USB-C ports on the iMac support USB 3.1 gen 2 (or 3.2 gen 2?) and connect at 10Gbps which is important for any SSD, even a SATA3 version.
The USB-A ports on my M1 Mini are only 5Gbps, which means the only way I can get 10Gbps connections on some external SSDs is to connect via a TB3 dock, and in any case it's faster to connect via a dock (TB3->USB3) than to connect directly to the TB3/USB4 port on the Mac even when both connect at 10Gbps. Oy vey!
The higher end M1 iMacs have two USB 3.1 gen 2 (or USB 3.2 gen 2x1 - same thing) ports from an ASMedia ASM3142 USB controller. The ASMedia ASM3142 will have faster USB (whether its 10 Gbps or 5 Gbps) than the the M1 Thunderbolt ports. Since the ASM3142 is not connected via Thunderbolt, it will probably be faster than Thunderbolt connected USB controllers as well (less latency without Thunderbolt connection). The ASM3142 is a PCIe device - I wonder if the PCIe setup of the M1 iMac is similar to the other M1 Macs (described at https://forums.macrumors.com/threads/when-will-apple-silicon-support-pcie-5.2288812/post-29702990 )
Apple could have done like the M1 Mac mini and just had USB 3.0 ports. The lower end M1 iMacs don't have any extra USB (like the M1 MacBook Pro and Air). Maybe the USB 3.0 controller of the M1 only works for type A which is too large or deep for the M1 MacBook Pro/Air/iMac.
Apple could have done like the M1 Mac mini and just had USB 3.0 ports. The lower end M1 iMacs don't have any extra USB (like the M1 MacBook Pro and Air). Maybe the USB 3.0 controller of the M1 only works for type A which is too large or deep for the M1 MacBook Pro/Air/iMac.
Perhaps someone could check the system report on an M1 laptop and see how their keyboard, trackpad and (particularly) webcam are connected? On some older Macs, those things have certainly been connected via internal USB (my laptop is a bit too old too be significant, but my 2017 iMac has the webcam connected via internal USB).
My speculation was that the M1 Mini was heavily based on the laptops and the "extra" USB3.0 ports were made possible because it doesn't have a keyboard, trackpad or webcam, freeing up a USB stream (possibly direct from the SoC?) The only M1 Mini teardown I can find doesn't list any USB 3 controller, just the TI CD3218B12 which (as far as I can tell) is just a USB-C port/PDC controller, which would be why it is also present on the iMac.
That the iMac would need a "proper" USB 3.1 controller would then follow - since it does have a webcam that might use a USB3 stream. Also goes some way to explain the apparently pointless distinction between 2 port and 4 port models if they're actually adding a controller.
There's so much we don't know about the M1 - for instance, Intel chipsets typically had a bunch of "universal" I/O lines which could be configured by the system builder as some permutation of PCIe, SSD-optimised PCIe, USB2/3, SATA etc. Is there some similar scheme for the M1? Are all the M1 chips in the new machines the same M1 (apart from RAM size and number of GPU cores) or are there more differences (...is the extra cooling on the 4-port iMac just for the extra GPU core...?)
Interesting theory, except the M1 Mac mini has a PCIe XHCI where the M1 MacBook Pro and M1 MacBook Air do not. The built-in keyboard and trackpad of the MacBook Air is definitely not USB:
Code:
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportBootloaderHIDDevice",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Device Management",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Device Management/AppleHIDTransportHIDDevice",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Device Management/AppleHIDTransportHIDDevice/IOHIDInterface",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Device Management/AppleHIDTransportHIDDevice/IOHIDInterface/AppleDeviceManagementHIDEventService",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Device Management/AppleHIDTransportHIDDevice/IOHIDInterface/AppleDeviceManagementHIDEventService/IOHIDEventServiceUserClient",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Keyboard / Boot",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Keyboard /AppleHIDTransportHIDDevice",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Keyboard /AppleHIDTransportHIDDevice/IOHIDInterface",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Keyboard /AppleHIDTransportHIDDevice/IOHIDInterface/AppleHIDKeyboardEventDriverV2",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Keyboard /AppleHIDTransportHIDDevice/IOHIDInterface/AppleHIDKeyboardEventDriverV2/IOHIDEventServiceUserClient",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Trackpad / Boot",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Trackpad /AppleHIDTransportHIDDevice",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Trackpad /AppleHIDTransportHIDDevice/IOHIDInterface",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Trackpad /AppleHIDTransportHIDDevice/IOHIDInterface/AppleMultitouchTrackpadHIDEventDriver",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Trackpad /AppleHIDTransportHIDDevice/IOHIDInterface/AppleMultitouchTrackpadHIDEventDriver/AppleMultitouchDevice",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Trackpad /AppleHIDTransportHIDDevice/IOHIDInterface/AppleMultitouchTrackpadHIDEventDriver/AppleMultitouchDevice/AppleMultitouchDeviceUserClient",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Actuator",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Actuator/AppleHIDTransportHIDDevice",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Actuator/AppleHIDTransportHIDDevice/IOHIDInterface",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Actuator/AppleHIDTransportHIDDevice/IOHIDInterface/AppleActuatorHIDEventDriver",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Actuator/AppleHIDTransportHIDDevice/IOHIDInterface/AppleActuatorHIDEventDriver/AppleActuatorDevice",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Actuator/AppleHIDTransportHIDDevice/IOHIDInterface/AppleActuatorHIDEventDriver/AppleActuatorDevice/AppleActuatorDeviceUserClient",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Accelerometer",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Accelerometer/AppleHIDTransportHIDDevice",
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/spi3@3510C000/AppleSPIMCController/ipd@0/AppleHIDTransportDeviceSPI/AppleHIDTransportProtocolHIDSPI/Accelerometer/AppleHIDTransportHIDDevice/IOHIDInterface",We don't know if the webcam is USB. Maybe it is and it is connected to the XHCI controller that the M1 Mac mini has. That would reduce the number of ports (without a hub) so Apple adds the ASM3142. But wait, the MacBook Air has a camera and it is not USB.
Code:
"/Root/J313AP/AppleARMPE/arm-io@10F00000/AppleT810xIO/isp@2A000000/AppleH13CamIn/AppleH13CamInUserClient"We don't know if the Ethernet in the power supply of the M1 iMac is PCIe (like the M1 Mac mini) or USB or Apple Silicon.
Good questions. The AppleT810xIO handles most/all I/O, including Thunderbolt. It has one AppleT8103PCIe for PCIe devices (one root port/two devices for MBA or MBP, three root ports/4 devices for MM) and a couple AppleT8103PCIeC - one for each Thunderbolt port to handle PCIe.
Ah. That just about wraps it up for that theory then...
The iFixit teardown found a BCM57762 on the M1 iMac logic board so its looking like PCIe. Which also suggests that the new mag connector is likely just acting as a hard-wired Ethernet cable ("Ethernet Over Power"?), so no interesting multi-function, or 10GbE power bricks in the near future...
I have a Cable Matters USB 3.1 SSD enclosure and it indeed only connects at 5 Gb/s. I inquired to them about it and they swore it should work. I pressed them and they finally tested it themselves and reported back to me that it is a limitation of the Apple hardware. 1) I was shocked they did not already know about this. 2) they also reported that it will work if you attach a TB dock to the Mac and then attach the drive to that. Yeah, they are really up to speed! pun intended.
So going with the idea of driving the enclosure from a TB dock, I got to thinking about if it would be possible to connect the drive to my Apple TB 27" display. It has a TB 2 port on the back. Yeah, I could purchase a TB4 dock, but I don't really want to add to the vast number of cables and hubs already attached to the Mini, not to mention the price of such a dock.
Anyone know if there is a TB2 -> USB 3.1 cable/adapter? I did some searching for this and really could not find anything.
Thanks in advance!
So going with the idea of driving the enclosure from a TB dock, I got to thinking about if it would be possible to connect the drive to my Apple TB 27" display. It has a TB 2 port on the back. Yeah, I could purchase a TB4 dock, but I don't really want to add to the vast number of cables and hubs already attached to the Mini, not to mention the price of such a dock.
Anyone know if there is a TB2 -> USB 3.1 cable/adapter? I did some searching for this and really could not find anything.
Thanks in advance!
A Thunderbolt 2 port can do Thunderbolt or DisplayPort. A Thunderbolt 2 port cannot do USB.
To connect USB to Thunderbolt 2, you need a Thunderbolt dock or Thunderbolt display. Thunderbolt 2 devices are usually limited to USB 5 Gbps. You could put a USB 10 Gbps PCIe card in a Thunderbolt 2 PCIe expansion box but in that case you might as well get a Thunderbolt 3 dock and an Apple Thunderbolt 3 to Thunderbolt 2 adapter if you are connecting the Thunderbolt 3 device to a Thunderbolt 2 host.
Can anyone with the 4-port M1 iMac check a Samsung T5 (or any 3.1 Gen 2 enclosure with SATA SSD) on the USB ports? Does it negotiate at 10 gb/s link speed?
As those 2 ports are using a dedicated PCI-e to USB 3.1 Gen 2 controller, my guess is that it should behave similar to Intel Macs and run at full speed.
Would be really helpful if someone can check and confirm. Cheers!
As those 2 ports are using a dedicated PCI-e to USB 3.1 Gen 2 controller, my guess is that it should behave similar to Intel Macs and run at full speed.
Would be really helpful if someone can check and confirm. Cheers!
Because I now have an M1 Mac and I was curious, here are some results.
Tested products:
The SSK enclosure does link at 10Gb/s when connected with a USB-C to USB-C cable on either port.
As for the speed verification.. Well, it is faster than USB 3.0 benchmarks, but not by leaps and bounds.
Blackmagic Disk Speed Test - 5GB
TechTool Pro 14 - 3GB and 1GB
Connect to the M1 Mac mini via USB A (3.0), the read is ~360 MB/s, write is ~370MB/s. In comparison, when the enclosure/drive was connected to my 2012 Mac mini, the read and write rates were ~495MB/s to 520MB/s.
I’m not sure why the write is generally a little faster on the M1 — especially because with Blackmagic the write value fluctuated massively as the test looped.
In conclusion, M1 Macs can establish a 10Gb/s USB link, however, the performance is underwhelming and less than with Intel Macs.
Why? Either bugs/inefficiencies with Big Sur’s I/O overall or with Apple’s arm64 versions of the drivers.
P.S. If there was already a post verifying 10Gb/s, this can be a confirmation.
Tested products:
SSK Aluminum M.2 NVME SATA SSD Enclosure Adapter, USB 3.2 Gen 2 (10 Gbps) to NVME PCI-E SATA M-Key/(B+M) Key Solid State Drive External Enclosure Support UASP Trim for NVME/SATA SSDs 2242/2260/2280
SSK Aluminum M.2 NVME SATA SSD Enclosure Adapter, USB 3.2 Gen 2 (10 Gbps) to NVME PCI-E SATA M-Key/(B+M) Key Solid State Drive External Enclosure Support UASP Trim for NVME/SATA SSDs 2242/2260/2280
www.amazon.com
Silicon Power 1TB - NVMe M.2 PCIe Gen3x4 2280 SSD (SP001TBP34A60M28)
Silicon Power 1TB - NVMe M.2 PCIe Gen3x4 2280 TLC SSD (SP001TBP34A60M28)
www.amazon.com
The SSK enclosure does link at 10Gb/s when connected with a USB-C to USB-C cable on either port.
As for the speed verification.. Well, it is faster than USB 3.0 benchmarks, but not by leaps and bounds.
Blackmagic Disk Speed Test - 5GB
TechTool Pro 14 - 3GB and 1GB
Connect to the M1 Mac mini via USB A (3.0), the read is ~360 MB/s, write is ~370MB/s. In comparison, when the enclosure/drive was connected to my 2012 Mac mini, the read and write rates were ~495MB/s to 520MB/s.
I’m not sure why the write is generally a little faster on the M1 — especially because with Blackmagic the write value fluctuated massively as the test looped.
In conclusion, M1 Macs can establish a 10Gb/s USB link, however, the performance is underwhelming and less than with Intel Macs.
Why? Either bugs/inefficiencies with Big Sur’s I/O overall or with Apple’s arm64 versions of the drivers.
P.S. If there was already a post verifying 10Gb/s, this can be a confirmation.
The 2012 Mac mini only has USB 3.0 which is 5 Gbps using 8b/10b encoding which allows 4 Gbps or 500 MB/s of data max. There's no way it can be higher than 500 MB/s.
Your 10 Gbps speeds on the M1 Mac are super low even for M1 Macs. If you had a Thunderbolt dock and maybe a display connected to it, then it may speed up the performance according to some tests done by OWC.
https://forums.macrumors.com/thread...lso-definitely-not-usb4.2269777/post-29935635
The performance would probably be greater than that if you connected the USB device to the Thunderbolt 3 dock.
Evidently, I was indeed mistaken on the recollection. I retested and the values were ~395MB/s write and ~420MB/s read. On the silver lining, that does put the M1 USB 3.0 performance closer to average, although, one would hope for about equal.
I know speeds never reach the theoretical/designed max (unless they’re artificially limited). In this context, don’t expect a USB 3.0 transfer speed to hit or come very close to 625MB/s. Anyway, I’m curious, what is “8b/10b”?
Not too surprising, the enclosure was on the “budget” end. I do have another NVMe enclosure...
RIITOP NVMe Enclosure M.2 NVMe to USB C Case External USB 3.1 to PCI-e (M Key) NVMe 2280 SSD Case Reader with 2 Cable
NVME to USB C Gen2 10Gb Adapter Enclosure RIITOP PCIe M.2 NVMe SSD to USB 3.1 Type C Converter Case in Black (Thunderbolt 3 Compatible) Features: Supports TRIM to the SSD. Integrates with USB Type-C multiplexer& configuration channel(CC)logic. Complies with PCI Express Base Specification Revision...
www.amazon.com
Any who… I’m not doing a lot of I/O intensive stuff, so it’s not a big deal, just disappointing that a PCIe SSD is limited to SATA SSD performance. Hopefully, Apple eventually gets I/O sorted out/optimized.
Thanks for the info.
8b/10b is the encoding of data on the wire used by USB 3.0 (and many other things such as PCIe, SATA, and DisplayPort).Evidently, I was indeed mistaken on the recollection. I retested and the values were ~395MB/s write and ~420MB/s read. On the silver lining, that does put the M1 USB 3.0 performance closer to average, although, one would hope for about equal.
I know speeds never reach the theoretical/designed max (unless they’re artificially limited). In this context, don’t expect a USB 3.0 transfer speed to hit or come very close to 625MB/s. Anyway, I’m curious, what is “8b/10b”?
Not too surprising, the enclosure was on the “budget” end. I do have another NVMe enclosure...
…But it’s currently coupled with the factory drive from my HP Envy as extra storage for my Xbox One.RIITOP NVMe Enclosure M.2 NVMe to USB C Case External USB 3.1 to PCI-e (M Key) NVMe 2280 SSD Case Reader with 2 Cable
NVME to USB C Gen2 10Gb Adapter Enclosure RIITOP PCIe M.2 NVMe SSD to USB 3.1 Type C Converter Case in Black (Thunderbolt 3 Compatible) Features: Supports TRIM to the SSD. Integrates with USB Type-C multiplexer& configuration channel(CC)logic. Complies with PCI Express Base Specification Revision...www.amazon.com
Any who… I’m not doing a lot of I/O intensive stuff, so it’s not a big deal, just disappointing that a PCIe SSD is limited to SATA SSD performance. Hopefully, Apple eventually gets I/O sorted out/optimized.
Thanks for the info.
https://en.wikipedia.org/wiki/8b/10b_encoding
5Gbps * 8b/10b = 4 Gbps = 500 MB/s. So the fastest speed you can expect from USB 3.0 is less than 500 MB/s (approximately 460 MB/s)
Faster protocols such as Thunderbolt, PCIe 3.0, USB 3.1 gen 2, USB4, use more efficient encodings such as 64b/66b, 128b/130b, 128b/132b
https://en.wikipedia.org/wiki/64b/66b_encoding
When testing maximum bandwidth, it's probably best to use a benchmark that returns the highest numbers - I believe AmorphousDiskMark.app does that.
For testing shared bandwidth, ATTO Disk Benchmark.app can test multiple devices at the same time without having to create a RAID 0.
As for USB to NVMe enclosures - I don't think there are any that can't do 1000 MB/s no matter how cheap. I have these ones:
https://plugable.com/products/usbc-nvme
https://www.newegg.ca/orico-tcm2-c3-bk-bp-enclosure/p/0VN-0003-001D8
Has anyone reported this to Apple tech support? I imagining so, but have not actually heard what Apple's response was.
Just got a new Orico 10Gbps NVME case with the RTL9210. I'm getting the full speed 10Gbps on battery on the MBA M1. NVME disk is a 500GB WDSN731 which is the OEM version of the SN750.
The other Orico case I have is 10Gbps but with the JMicron chipset and it's limited to 5Gbps unless I plug it into a TB3 dock.
The other Orico case I have is 10Gbps but with the JMicron chipset and it's limited to 5Gbps unless I plug it into a TB3 dock.