Sabrent NT-SS5G Review USB to 5GbE NIC

[Rohit Kumar]

In our Sabrent NT-SS5G review, we see how this USB 3.1 Gen1 to 5GbE adapter based on the Marvell AQC111U controller performs and compares

The post Sabrent NT-SS5G Review USB to 5GbE NIC appeared first on ServeTheHome.

Continue reading...

 

[geppi]

This adapter does show some strange performance on my side.
Whatever I do I can't get more than about 150 MB/s average throughput when the adapter is working as the sending side of a connection.

I've tested with iperf3 and it gives me a stable throughput of about 3.45 Gb/s when the machine with the Sabrent adapter is acting as the iperf3 server side, i.e. the receiving end.
However, when I switch the connection around and have the Sabrent machine act as the client, i.e. the adapter is sending, the throughput varies between 1.9 Gb/s and about 500 Mb/s with an average of about 1.2 Gb/s.

This is in line with the results of copying files from one machine to another which shows a stable transfer of about 420 MB/s when the machine with the Sabrent adapter is the target and an extremely fluctuating throughput of in average 150 MB/s when it is the source.


I've tested this with the Sabrent adapter connected to a Zyxel XS1930-10 Multi-Gigabit-Switch as well as directly connected to another PC with an Asus XG-C100C network adapter.
Behind the switch the Sabrent adapter did talk to machines with Intel 10GigE network adapters. All machines are perfectly capable to saturate a 10GigE link and transfer files with about 980 MB/s in both directions between each other.

I did test the adapter on two different PCs with various driver parameter settings. No matter what I did it was always the same picture.

It sounds like a stupid question, but did you guys test the adapter in both directions ?

 

[eva2000]

In our Sabrent NT-SS5G review, we see how this USB 3.1 Gen1 to 5GbE adapter based on the Marvell AQC111U controller performs and compares

The post Sabrent NT-SS5G Review USB to 5GbE NIC appeared first on ServeTheHome.

Continue reading...

Thanks for the review. As the 5Gbe units overhead seem larger than 2.5Gbe units, maybe manufacturers are also artificially capping speeds on 5Gbe as a way of thermal throttling management?

It sounds like a stupid question, but did you guys test the adapter in both directions ?

That is a interesting question as I was looking at these USB to 2.5Gbe/5Gbe adapters as a way of transferring data between 2 laptops so both sides would be using the adapters.

 

[geppi]

@Rohit Kumar It would be really helpful for me to know if your performance tests did also cover the NT-SS5G working as the sending device. If yes and the results were as published it would be a strong indication for me that the one I received has a hardware defect in the transmitter electronics.

 

[geppi]

OK, I can now exclude a hardware issue after receiving a replacement adapter and both show the same problem as described above.

However, when further investigating I discovered a real show-stopper for this adapter, at least with the current Windows 10 driver.

When I received the replacement adapter I initially connected it peer to peer with an Asus XG-C100C network adapter in my desktop PC and left the driver parameter settings on their default values, i.e. no jumbo frames and no buffer tweaking. In receiving mode the adapter showed about 3.2 Gb/s which didn't surprise me based on my earlier tests and the fact that the settings were not optimized. What on the other hand did surprise me was that I did get about 2.1 Gb/s when the adapter was in sending mode. This throughput was even stable and did not vary as much as in my earlier tests.
After enabling jumbo frames and optimization of the ring buffer settings I was able to bump this up to the 3.45 Gb/s in receiving mode and about 3 Gb/s in sending mode. So what was wrong with my earlier tests ?

The explanation turned up short after I enabled "Priority & VLAN" on both adapters. With VLAN enabled and a VLAN ID other than "0" the sending performance did exactly show the same picture as in my initial tests, i.e. very unstable with low values in the Mb/s range and maximum values of about 1.9 Gb/s and an average of about 1.2 Gb/s. The receiving performance was not impacted.

To assure that I hadn't screwed up something with my tweaked parameter settings I repeated the test with the plain vanilla driver settings and it delivered a similar result. As long as no VLAN ID was set the performance was OK and configuring VLAN with an ID other than "0" brought the sending performance down.

But that's not the end to the story. What's even worse is that no matter what VLAN ID you set on the Sabrent adapter, it doesn't matter !
You can still ping between the to computers even when you set them to different VLAN IDs.
So the Sabrent adapter doesn't honor the VLAN ID setting !
Well that's not exactly true, it honors it with a lower performance.

I did open a ticket with Sabrent and it will be interesting to see how that moves on.

Again I would be very interested to understand what tests were performed at STH on this adapter ?
Since this looks to me very much like a driver issue I would also expect that the StarTech adapter, which also uses the Marvell driver, shows the same behaviour.

 

[eva2000]

@geppi thanks for sharing your findings. Keep us up to date!

 

[geppi]

@geppi thanks for sharing your findings. Keep us up to date!

For your use case of a peer to peer connection between 2 laptops the VLAN issue is probably not that important. However, it shows that the current driver is buggy and you never know what other crocodiles are lurking in other corners.

Also, I still find it not convincing to see about 500 Mb/s less performance when sending compared to the receiver mode when the VLAN issue doesn't hit. But maybe there is a mistake on my side which I currently don't see. That's why it would really be interesting to see other performance tests.

 

[tigweld0101]

I usually tag at the switch port so I haven't seen the vlan thing. I've seen consistent 3.2-3.5gbps from this using it as the iperf3 client and server. That's what I saw STH publish.

I'm just using the marvell driver not the sabrent one.

 

[geppi]

I usually tag at the switch port so I haven't seen the vlan thing. I've seen consistent 3.2-3.5gbps from this using it as the iperf3 client and server. That's what I saw STH publish.

I'm just using the marvell driver not the sabrent one.

They're identical.

Thanks for your input.

 

[geppi]

@tigweld0101 Did you change any of the driver settings to get the send performance above 3 Gb/s or was it just out of the box ?

 

[mikew4]

Signed up to bring this one back and wanted to see what folks' test results are with the Sabrent adapter.

Key issue: Not seeing the speeds I expect when using the USB A to USB C cable for the Sabrent.
I do see speeds that I expect when using the USB C to USB C cable.

With the USB A cable, I found the following:
-Link speed may or may not work as expected (5G)
-Changing driver settings had a mild impact in overall test results
-Actual speed throughput whether checking via file transfer or via iperf3 was sorely disappointing using the USB A cable

Here's what I found when using the USB A cable:

System	Driver	Max Link Speed	iperf3 speeds	File transfer speeds
Windows 10 20H2 on a 2019 HP Spectre x360 15"	Latest from Marvell site	1Gb	350Mb/s - 450Mb/s (settings dependent)	N/A
Windows 10 1909 on a Dell Precision M4800	Driver from the dongle itself	5Gb	500Mb/s - 580Mb/s (settings dependent)	N/A
Synology NAS 1813+	Open source package	5Gb	N/A	950Mb/s

I was wondering if anyone else's results differed so wildly based on the cable used or not.

My real use case is for the Synology via a USB A cable, so if what I'm seeing is consistent with other Sabrent adapter users, I'll need to try another adapter to get the results I'm looking for.

Any feedback that others could provide additional context to my testing would be quite helpful.

Thanks!