I started working on USB in 1995, starting with the world’s first BIOS that supported USB Keyboards and Mice while at Award Software. After a departure into embedded systems software for real-time operating systems, I returned to USB IP cores and software at inSilicon, one of the leading suppliers of USB IP. In 2002, inSilicon was acquired by Synopsys and I’ve been here since. I also served as Chairman of the USB On-The-Go Working Group for the USB Implementers Forum from 2004-2006. USB 3.0 SSD Benchmark – First Published Results
Posted by Eric Huang on November 12th, 2009
Reading through the first published benchmark testing of a USB 3.0 SSD, you can look at it 2 ways.
1) It’s slower than the promised 10 times USB 2.0 or
2) It’s 3 times to 7 times faster than USB 2.0 and almost as fast as direct SATA 3 Gbps connection.
The way it works, the SSD actually has a SATA 3 Gigabit per second (Gbps) connection. To get to USB 3.0, it uses a SATA to USB 3.0 chip. (I’d really like to know which one.). This means the fastest the device can go is 3 Gbps. This means it can’t ever get to 4 Gbps.
(Remember, electrical signaling for USB 3.0 is at 5 Gbps, and with 8b10b encoding it drops by to a 4 Gbps maximum theoretical limit).
Looking at the bench mark, you will see USB 3.0 transfers within 10% or less of the SATA transfers. In SATA mode, it’s actually connected directly to the PC Chipset SATA connector. So there is some overhead translating from SATA to USB.
So my assessment is, this is a pretty good start for USB 3.0 Devices.
Full link to the report is here
Also, the pre-brined Turkey’s are in at Trader Joe’s so clear some space in your fridge.
RSS