[arin-ppml] Policy Proposal: A Modest Proposal for an Alternate IPv6 Allocation Process
George, Wes E [NTK]
Wesley.E.George at sprint.com
Fri Jun 5 15:54:44 EDT 2009
> Ted Mittelstaedt wrote:
<snip> both those companies continued to select esoteric and expensive parts to use so as to keep their unit
costs high, they would just be opening the door to the Linux-based companies (like Imagestream) who are waiting in the wings. Cisco is well aware of this, it is after all no surprise that they recently reorganized their Linksys purchase and pulled all the high-end linux-based routers (ie: rv082, rvs1000, etc.) out of that brand.
And I suppose the 100mpg carburetor would have been right around the corner if Big Oil hadn't killed it too? There's a name for esoteric and expensive parts. It's called cutting edge. Someone has to invent the stuff and implement it before it commoditizes and gets cheap. I'm not going to say that the router vendors in question are not milking more profit out of some of their hardware, but you're going way far out into tinfoil-hat land with a lot of this thread so far.
20 years from today, we will have hardware so fast that everything
implemented in a hardware asic on a router today will be easily
possible to duplicate in software.
Yes, and in 20 years we'll be fondly looking back on the bad old days when data rates were still measured in Gigabits per second and you had to carry your computer around (gasp). Every person on the planet will have a 1 Terabit/sec wireless connection to the chip in your head, meaning that the core will be handling amounts of data that we probably don't have an SI prefix for yet. And sorry, that will still probably require hardware-based acceleration. It's simply not possible to get the same performance out of general purpose hardware and software running on top of it than by purpose built hardware. That may change in the future, but what I'm concerned with is not the size of the routing table in 20 years. I'm concerned that the routing table will grow faster than our ability to innovate new technology to deal with it, or worse, that it will continue to cost too much, and require upgrades every few months instead of every few years. You want to complain about the cost of memory? The unobtanium technology necessary to keep up with a routing table growing exponentially will make that pale by comparison.
Please stop trying to discuss core router architecture on this list. You want to get involved in the discussion about what the internet routing infrastructure might look like in 20 years? http://www.irtf.org/charter?gtype=rg&group=rrg is the perfect list for that.
This e-mail may contain Sprint Nextel Company proprietary information intended for the sole use of the recipient(s). Any use by others is prohibited. If you are not the intended recipient, please contact the sender and delete all copies of the message.
More information about the ARIN-PPML