[kwlug-disc] USB 2.0 vs. 3.0 on Linux

[William Park]

On Mon, Sep 22, 2014 at 08:16:06PM -0400, Khalid Baheyeldin wrote:
> On Mon, Sep 22, 2014 at 7:59 PM, William Park <opengeometry at yahoo.ca> wrote:
> 
> > - What kind of test?
> >
> 
> Copying around 100GB worth of files, most of them multi-gigabyte in size,
> from one disk to the other. Using rsync, while the files do not exist on
> the target
> 
> rsync -av /mnt/disk1/ /mnt/disk2

You have to factor in CPU/memory/chipset here.  Unavoidable, I guess,
since you're comparing USB2 vs USB3.

> 
> 
> > - Sequential, random?
> >
> 
> Sequential.

Then, 20MB/s USB2 and 50MB/s USB3 sound right.

> 
> 
> > - DC or USB powered?
> >
> 
> USB powered. Both disks are portable (one Toshiba Canvio, the other
> Seagate Expansion.

Other factors...

    - Most importantly, how much power you can get from USB port and how
      fast.  The 5V pins on USB port is not 5V rail from power supply.
      Rather, it's "negotiated" with USB device.  

    - To save power, they use "standby" and "sleeping" features
      liberally.  You see the difference between DC powered and USB
      powered, even if disk is spinning.  

    - Even at max, USB2/USB3 can only supply 0.5A/0.9A respectively,
      which means 2.5W/4.5W respectively.  Normal harddisk is around
      10W.

> 
> 
> > - Intel, AMD? -- Intel 3Gbps chipset is better than AMD 6Gbps.
> 
> 
> This is an AMD machine.
> 
> 
> > You see the difference under heavy load, like 4 VMs in
> > Hyper-V as I'm doing now.
> 
> Bare metal in my case, no virtualization, and no other load on the machine.
> 
> Seems there is a partial answer here:
> 
> http://www.macworld.com/article/2039427/how-fast-is-usb-3-0-really-.html
> 
> Spinning disks can only do 114MB/s (7200 RPM that is). SSDs can do 200 MB/s.

For sequential write, I usually do

    - dd if=/dev/zero of=/dev/sdX bs=32M oflag=direct
    - sg_dd if=/dev/zero of=/dev/sdX blk_sgio=1 time=1

After a shot of expresso or two, send USR1 to the process to get some
stats.
-- 
William