No announcement yet.


  • Filter
  • Time
  • Show
Clear All
new posts

  • USB3Test

    Hi, I am a recent buyer of the USB3Test plug and have been firing it up on various cables to check out the performance on my laptop. Two questions:
    1. When I select Connection type to Superspeed (5Gbps) or Highspeed (480Mbps) no issues, the USB3 plug responds correctly. When I select Fullspeed (12Mbps) the USB3 plug responds with Highspeed setting. Not that it is a problem, it just surprised me. Is this what you are expecting to occur?
    2. When I perform Superspeed benchmark tests with only read or only write selected I get actual throughput of 3.1Gbps, totally what I expect. But when I do simultaneous read and write, both read and write drop down to 2Gbps each. This was not expected. Is this a limitation of the laptop hardware, USB3 loopback plug? I was wanting to check that 3+ Gbps is maintained over read and write. Is the fact that it drops down to 2Gbps showing an issue with the laptop/PC or cable or a limitation of the Passmark loopback plug? Not an issue, I still love the Passmark loopback plug and the USB3Test s/w for the price. I just need an honest answer on where the limitation might live.

  • #2
    1. In some testing today, we saw the same thing. Which is strange as I was fairly sure it was working at 12Mbps at some point in the past with at least some motherboards. I wonder if newer host controllers are automatically re-negotiating the link speed.

    2. This is more or less in line with what we see. We haven't had a deep look into the cause. But I suspect it is because sending data in one direction automatically results in packets coming back from the other (ACK - acknowledgment packets). We only measure the amount of user data being transferred. So the bytes in these low level protocol packets are not included in the transfer rate bps. I also suspect these protocol packets interrupt the BULK data transfer. Leading to a certain amount of inefficiency with simultaneous bi-directional data transmission.


    • #3
      Thanks for confirming that your testing replicates my observations in both cases. Yes, I only want to see the user data throughput rate. That is exactly what I am expecting.
      Thanks for the great simple products that are easy to use and great technical support. Keep it going.