Provided that the exponentially more transcoding that they'll try to do is equal to or less than the exponentially Moore increase in processing power over time*, this solution comes with an in-built growth path. --Dave * I'm aware that that's not quite what Mr. Moore said. On Thu, Aug 27, 2015 at 2:16 PM, Alex Balashov <abalashov at evaristesys.com> wrote:

Strongly agree with Simon that transcoding several thousand concurrent channels on a multicore box is highly possible in software, using FOSS, and that it would most likely yield the best per-port economics, learning curve and ongoing maintenance notwithstanding.

However, Brooks' stated main criteria isn't cost so much as physical density and power footprint. With that in mind, I'm not sure if a single 1U server can reasonably do 5000 channels in software _and_ provide a future growth path. Companies that realise they can do X much more economically than they thought possible tend to do exponentially more X[1], so I could see a home-grown solution branching out into a 3-5 box setup easily. As I understood it, that's something Brooks is looking to avoid here, primarily due to rackspace and power constraints.

-- Alex

[1] In the style of https://en.wikipedia.org/wiki/Jevons_paradox

-- Alex Balashov | Principal | Evariste Systems LLC 303 Perimeter Center North, Suite 300 Atlanta, GA 30346 United States

Tel: +1-800-250-5920 (toll-free) / +1-678-954-0671 (direct) Web: http://www.evaristesys.com/, http://www.csrpswitch.com/

_______________________________________________ VoiceOps mailing list VoiceOps at voiceops.org https://puck.nether.net/mailman/listinfo/voiceops

-- David Knell, Director, TelNG T: +44 1223 797979 / +1 970-315-4721 W: http://www.telng.com H: http://www.daveknell.com