USB 3.2 for embedded systems

I have always found USB a fascinating technology. It is a wonderful example of how something can have great complexity inside, which is used to render the thing very easy to use. I have written and spoken about USB many times over the years – here, for example. So, I was quite interested when I saw that USB 3.2 was coming along …

When USB first appeared, back in the 1990s, it was very exciting. At a stroke, it made the lives of most computer users a bit easier. Installing new peripherals became a true plug ’n’ play experience. The data transfer speeds were modest, but it was a start. In 2000 USB 2.0 came along. This offered a significant improvement in data transfer rates [to 480Mbps], which was very welcome. Eight years later, USB 3.0 introduced a radical architectural change to USB to provide much better speeds again [5Gbps], while maintaining quite good backwards compatibility.

When USB 3.1 was announced not so long ago, there was another modest speed increase [to 10Gbps] and a very significant increase in the power delivery capacity of USB. At the same time, the small, reversible [yes!] Type-C connector was introduced. This paved the way to the use of USB as a power supply for laptops etc.

Expectations for USB 3.2 were not great, as it is only a “.1” update to the spec. And it is fair to say that the enhancements are incremental. When USB 3.0 was introduced, it added 4 extra data wires to the 2 that were used previously. These 4 wires facilitated simultaneous bi-directional data transfer at “SuperSpeed”. USB 3.1 increased this speed a little [10Gbps]. USB 3.2 provides the option to use the two channels in parallel in the same direction, thus effectively doubling the speed again [to 20Gbps].