[tor-talk] Tor relay on small and cheap devices

[Roger Dingledine]

On Sun, Jan 20, 2013 at 04:28:13PM -0800, Mike Perry wrote:
>  it's unlikely that the
> network could support enough of these tiny relays to actually make any
> substantial capacity difference, and they may actually harm overall
> performance rather than help it.

See also
https://trac.torproject.org/projects/tor/ticket/1854
where we're trying to work out enough measurement/evaluation/simulation
infrastructure to answer the question "If we drop all the relays with
capacity less than X, total network capacity goes down but speed goes
up. What does that look like for various values of X and various network
loads?"

If you want to go even deeper into the research, check out these recent
tech reports that explore splitting circuits into multiple paths to take
better advantage of smaller relays:
http://www.cacr.math.uwaterloo.ca/techreports/2011/cacr2011-29.pdf
http://www.cs.bu.edu/techreports/pdf/2012-013-tor-assist.pdf

--Roger