Re: [Manycouches] Daniel presentation @ 113 - "carbon emissions considerations"

[John Scudder <[e-mail]>]

To date, efforts I’ve seen that have the explicit goal of “we’re going to minimize carbon impact of <networking thing X>” run up against the fact that the industry has long had financial incentives to minimize power consumption and heat removal requirements, because those represent significant costs for anyone doing computing at scale. 

Never say never, but I’m not expecting to find any low-hanging fruit that hasn’t been harvested already.

—John

> On Apr 12, 2022, at 10:30 AM, Joel M. Halpern <[e-mail]> wrote:
> 
> While it would be nice to think we can redesign our protocols to be more
> cooperative with the environment, it seems clear that we do not have a
> clue how to do so.  Putting that on the table even as a  future seems to
> be misleading.  One of the other SDOs I work with puts an "energy
> considerations" section in all their documents.  There are a few where
> that may make sense (when they deal with laser power levels, etc).  But
> in general it is a meaningless section serving to pretend to address an
> issue.
> 
> Yours,
> Joel
> 
> On 4/12/2022 5:56 AM, Jay Daley wrote:
> ...
>> Finally, I have a strong suspicion that we are only talking about doing
>> our part by reducing flying because it’s so visible a contributor not
>> because it has the most potential.  Like Martin, it seems obvious to me
>> that the biggest potential is in the protocols themselves, but that of
>> course is hard to measure, hard to understand how design affects
>> emissions, and hard to get people to look at.  Having said that, I can
>> certainly see a point when RFCs have an "Carbon emissions
>> considerations" section or something broader.
>> 
>> Jay
>