Every Thunderbolt 3 port can also work as a regular USB-C port, but I haven't seen a USB hub that actually taps into Thunderbolt 3 rather than regular USB 3.1 Gen 1 or 2.
Actually it turns out there is at least one Thunderbolt 3-based USB hub, which I’ve linked below. Its advantage over traditional hubs is not only that it connects back to the system over Thunderbolt 3 rather than USB 3.1 Gen 1 or 2, but also that it implements multiple USB controller chips within the hub itself, which means each connected device gets a dedicated USB 3.1 Gen 1 or 2 interface with the hub. Practically every hub on the market (and most PCs, for that matter) only have a single controller chip that all of their ports connect to, which means the maximum total bandwidth across all ports is only USB 3.1 Gen 1 or 2, whichever the controller supports. But as you’ll see, this hub is QUITE a bit more expensive as a result: https://www.startech.com/Cards-Adapters/USB-3.0/Hubs/thunderbolt-3-usb-3-1-hub~TB33A1C
UPDATE: Based on the product description mentioning 20 Gbps aggregate throughput and the wording in the manual, it sounds like that product has individual USB 3.1 Gen 1 controllers for the two 3.1 Gen 1 ports, and then a single USB 3.1 Gen 2 controller shared between the two USB 3.1 Gen 2 ports. That was likely a cost-driven decision given that there are relatively few USB 3.1 Gen 2 peripherals on the market and even fewer that would fully utilize its 10 Gbps of bandwidth for a sustained period of time. Even a SATA-based SSD wouldn't manage that (although an NVMe SSD would be significantly bottlenecked there.)
@gracieallen2 sorry for the triple posts here, but I just wanted to check one detail. In the multi-device connection scenarios where you say throughput is bad, does that apply when both devices are actively in use, e.g. transferring files, or does performance suffer when multiple devices are connected even if only one of them is actively in use? If it's the former, the fact that both of your scenarios mention an SSD is probably the key. There's a good chance that the SSD's USB enclosure only supports USB 3.1 Gen 1 (5 Gbps), and modern SATA SSDs can completely saturate that amount of bandwidth, leaving nothing for another device. Even if your SSD enclosure supports 3.1 Gen 2 (10 Gbps), I believe the USB-A ("regular USB") ports on that system only support 3.1 Gen 1, so that extra bandwidth wouldn't be available. And even if it were, if your OTHER devices only support 3.1 Gen 1, then I don't know if they'd be able to tap into the "leftover" 3.1 Gen 2 bandwidth that the SSD isn't using. But either way, if that’s your problem scenario, then the hub I linked above should make a significant difference. Also note that USB 3.1 is full duplex, which means that they can provide either 5 or 10 Gbps in each direction simultaneously -- so for example, in theory you should be able to copy data FROM your SSD to your PC and simultaneously copy data TO your spinning hard drive without those activities affecting each other's performance.
If on the other hand you're seeing bad performance with multiple device connected even when you're only actively using one, then I'm not sure what to suggest other than updating your system's BIOS/firmware, because that definitely shouldn't be happening at least on powered hubs. With unpowered hubs, you could end up with bad performance due to power requirements being too high for the system to satisfy. But if that’s the issue you’re seeing with multiple powered hubs, then there’s very little chance the hub I linked above would help.
Thanks for the replies... Yes, all the devices I'm connecting are, I suspect Gen 1. For example, I typical day has the Logitech receiver for the keyboard and trackball in one USB port (it works DREAFULLY when plugged into any of the hubs, regardless of whatever else is going on). In another USB port is an external HD, 7200 rpm with the images on it. It likewise has problems when plugged into any of the hubs - data transfers hang, then eventually throw an error. So when possible I try to limit the hub to things like a couple memory card readers and a flash drive or two.
On OCCASION I've plugged in the USB-powered USB and an external HD on the hub, and when I do BOTH suffer from very poor throughput and frequent hangs during data movement. I don't recall the exact error but it's something I was told is related to the way USB functions.
It gets to be a bigger problem when I'm in a situation where I have the camera tethered (1 USB), the USB SSD doing backups every 15 minutes (another USB), a flash drive with other data (3rd USB), the Logitech receiver (4th USB), a CD/DVD burner (5th USB) and a powered external HD connected (6th USB)... Since the laptop only HAS 3 USBs, and the camera tether does NOT like being on the hub, and the Logitech receiver does NOT like being on the hub, I have to put data devices on the hub...
I"ll have to keep an eye out and see what I can find for a Thunderbolt hub...
The Logitech receiver's performance could be due to the fact that it uses a 2.4 GHz wireless signal (not actual WiFi, but the same frequency range), and 2.4 GHz and USB 3.0 are known to interfere with each other. That's why some wireless receivers recommend plugging them into a USB 2.0 port if one is available. I suppose the receiver might be causing issues for actual USB 3.0 high-bandwidth devices when it's plugged into the hub, but it sounds like you've already experimented with that.
With the devices you're trying to run, unpowered hubs would be out of the question unless your 7200 RPM drive, SSD, and CD/DVD burner are all themselves wall-powered. But if for example you have only your SSD and 7200 RPM drive connected to a powered hub, and you're only transferring files to/from one of them, and that device still performs significantly worse than if you had only that one active device connected, I'm not sure how to account for that considering that you've already tried multiple powered hubs. And again, in that scenario I very much doubt a Thunderbolt hub would make a difference because it doesn't sound like bandwidth starvation between your hub and PC is your issue.
gracieallen2
23 Posts
0
May 12th, 2019 14:00
Anyone?
jphughan
9 Legend
•
14K Posts
0
May 12th, 2019 15:00
Every Thunderbolt 3 port can also work as a regular USB-C port, but I haven't seen a USB hub that actually taps into Thunderbolt 3 rather than regular USB 3.1 Gen 1 or 2.
jphughan
9 Legend
•
14K Posts
0
May 12th, 2019 17:00
Actually it turns out there is at least one Thunderbolt 3-based USB hub, which I’ve linked below. Its advantage over traditional hubs is not only that it connects back to the system over Thunderbolt 3 rather than USB 3.1 Gen 1 or 2, but also that it implements multiple USB controller chips within the hub itself, which means each connected device gets a dedicated USB 3.1 Gen 1 or 2 interface with the hub. Practically every hub on the market (and most PCs, for that matter) only have a single controller chip that all of their ports connect to, which means the maximum total bandwidth across all ports is only USB 3.1 Gen 1 or 2, whichever the controller supports. But as you’ll see, this hub is QUITE a bit more expensive as a result: https://www.startech.com/Cards-Adapters/USB-3.0/Hubs/thunderbolt-3-usb-3-1-hub~TB33A1C
UPDATE: Based on the product description mentioning 20 Gbps aggregate throughput and the wording in the manual, it sounds like that product has individual USB 3.1 Gen 1 controllers for the two 3.1 Gen 1 ports, and then a single USB 3.1 Gen 2 controller shared between the two USB 3.1 Gen 2 ports. That was likely a cost-driven decision given that there are relatively few USB 3.1 Gen 2 peripherals on the market and even fewer that would fully utilize its 10 Gbps of bandwidth for a sustained period of time. Even a SATA-based SSD wouldn't manage that (although an NVMe SSD would be significantly bottlenecked there.)
jphughan
9 Legend
•
14K Posts
0
May 12th, 2019 19:00
@gracieallen2 sorry for the triple posts here, but I just wanted to check one detail. In the multi-device connection scenarios where you say throughput is bad, does that apply when both devices are actively in use, e.g. transferring files, or does performance suffer when multiple devices are connected even if only one of them is actively in use? If it's the former, the fact that both of your scenarios mention an SSD is probably the key. There's a good chance that the SSD's USB enclosure only supports USB 3.1 Gen 1 (5 Gbps), and modern SATA SSDs can completely saturate that amount of bandwidth, leaving nothing for another device. Even if your SSD enclosure supports 3.1 Gen 2 (10 Gbps), I believe the USB-A ("regular USB") ports on that system only support 3.1 Gen 1, so that extra bandwidth wouldn't be available. And even if it were, if your OTHER devices only support 3.1 Gen 1, then I don't know if they'd be able to tap into the "leftover" 3.1 Gen 2 bandwidth that the SSD isn't using. But either way, if that’s your problem scenario, then the hub I linked above should make a significant difference. Also note that USB 3.1 is full duplex, which means that they can provide either 5 or 10 Gbps in each direction simultaneously -- so for example, in theory you should be able to copy data FROM your SSD to your PC and simultaneously copy data TO your spinning hard drive without those activities affecting each other's performance.
If on the other hand you're seeing bad performance with multiple device connected even when you're only actively using one, then I'm not sure what to suggest other than updating your system's BIOS/firmware, because that definitely shouldn't be happening at least on powered hubs. With unpowered hubs, you could end up with bad performance due to power requirements being too high for the system to satisfy. But if that’s the issue you’re seeing with multiple powered hubs, then there’s very little chance the hub I linked above would help.
gracieallen2
23 Posts
0
May 16th, 2019 05:00
Thanks for the replies... Yes, all the devices I'm connecting are, I suspect Gen 1. For example, I typical day has the Logitech receiver for the keyboard and trackball in one USB port (it works DREAFULLY when plugged into any of the hubs, regardless of whatever else is going on). In another USB port is an external HD, 7200 rpm with the images on it. It likewise has problems when plugged into any of the hubs - data transfers hang, then eventually throw an error. So when possible I try to limit the hub to things like a couple memory card readers and a flash drive or two.
On OCCASION I've plugged in the USB-powered USB and an external HD on the hub, and when I do BOTH suffer from very poor throughput and frequent hangs during data movement. I don't recall the exact error but it's something I was told is related to the way USB functions.
It gets to be a bigger problem when I'm in a situation where I have the camera tethered (1 USB), the USB SSD doing backups every 15 minutes (another USB), a flash drive with other data (3rd USB), the Logitech receiver (4th USB), a CD/DVD burner (5th USB) and a powered external HD connected (6th USB)... Since the laptop only HAS 3 USBs, and the camera tether does NOT like being on the hub, and the Logitech receiver does NOT like being on the hub, I have to put data devices on the hub...
I"ll have to keep an eye out and see what I can find for a Thunderbolt hub...
jphughan
9 Legend
•
14K Posts
0
May 16th, 2019 08:00
The Logitech receiver's performance could be due to the fact that it uses a 2.4 GHz wireless signal (not actual WiFi, but the same frequency range), and 2.4 GHz and USB 3.0 are known to interfere with each other. That's why some wireless receivers recommend plugging them into a USB 2.0 port if one is available. I suppose the receiver might be causing issues for actual USB 3.0 high-bandwidth devices when it's plugged into the hub, but it sounds like you've already experimented with that.
With the devices you're trying to run, unpowered hubs would be out of the question unless your 7200 RPM drive, SSD, and CD/DVD burner are all themselves wall-powered. But if for example you have only your SSD and 7200 RPM drive connected to a powered hub, and you're only transferring files to/from one of them, and that device still performs significantly worse than if you had only that one active device connected, I'm not sure how to account for that considering that you've already tried multiple powered hubs. And again, in that scenario I very much doubt a Thunderbolt hub would make a difference because it doesn't sound like bandwidth starvation between your hub and PC is your issue.