►
From YouTube: IAB workshop on Environmental Impact of Internet Applications and Systems 1: The Big Picture
Description
IAB workshop on Environmental Impact of Internet Applications and Systems (E-Impact) Session 1: The Big Picture
Workshop webpage: https://www.iab.org/activities/workshops/e-impact/
A
A
B
Are
being
recorded?
Yes,.
A
Yes,
good
and
the
record
is,
will
actually
be
published
in
YouTube
later
your
position
papers
for
those
submitted,
the
position
paper,
you
you,
they
have
been
made
public
on
the
Workshop
web
page,
and
this
is,
of
course,
a
professional
meeting.
So
we
expect
everybody
to
behave
in
a
professional
manner.
Any
kind
of
harassment
is
not
tolerated
and
I
do
want
to
say
that
this
is
like.
If
you
used
to
ITF
meetings,
then
this
is
maybe
a
little
bit
more
diverse
group
of
people.
A
So
not
everybody
has
the
same
background
and
it's
useful
to
remember
when
we
are
discussing
so
please
be
polite,
obviously,
but
also
explain
your
Viewpoint
in
an
understandable
Manner
and,
let's
all
learn
from
each
other's
viewpoints.
So
that's
the
reason
we
actually
are
here
and
today's
at
the
end
of
this
is
the
welcome
session
we're
going
to
cover
some
practicalities.
Why
we
are
here.
We
can
talk
about
some
big
picture
issues,
interaction
elsewhere.
How
this
relates
to
societal
issues
and
so
forth.
A
You
can
see
the
agenda
on
screen,
I,
hope
and
I'm
going
to
start
with
well
doing
this,
and
also
some
of
the
big
picture
issues
and
then
we'll
have
Taurus
Eve
and
visna,
and
it's
the
same
pattern
for
all
of
these
sessions.
We
actually
have
four
sessions,
so
this
is
the
first
of
the
four.
A
So
we
will
have
not.
Everybody
is
presenting
like
their
position
paper
or
most
people
aren't,
but
we
will
have
some
hopefully
exciting
presentations
at
the
beginning
to
get
us
started
and
then
we
can
discuss
and
the
discussion
is
actually
the
maybe
the
main
goal.
Perhaps
we
can
get
to
some
conclusions,
even
in
some
some
cases
or
at
least
Identify
some
issues.
A
Moving
on
so
the
upcoming
sessions
this
week,
Thursday
and
Friday
there's
two
sessions.
One
is
on
what
we
understand
about
the
situation.
So
that's
about
things
like
how
much
is
the
internet
actually
using,
for
instance,
energy
or
producing
carbon,
but
also
other
other
things,
and
what
what
do
we
don't
understand?
Can
we
do
the
whole
breakdown
of?
Where
is
the?
Where
is
the
impact
and
what
is
causing
it,
and
and
why
and
so
on
and
I?
A
Don't
actually
think
that
we
understand
everything
so
clearly,
there's
the
show
for
discussion
and
then
on
the
Friday
assessor,
we'll
talk
about
the
improvements
of
different
categories,
we'll
have
some
talks
on
specific
areas
of
technology
to
get
us
started
and
there's
no
discussion
and
then
on
the
next
Monday
exactly
a
week
from
now.
All
the
sessions
are
by
the
way.
At
the
same
time,
we
will
talk
about
what
the
conclusions
might
be
so
reporting
back
or
reflecting
back
a
little
bit
from
the
earlier
sessions
and
look
at
specific
areas.
A
Where
are
we
on
that
and
what
should
be
done
next
and
then
some
discussion
and
talking
about
how
we
go
about
reporting
and
so
forth,
I
by
the
way
apologize
that
it's
been
difficult
to
schedule
these
sessions
is,
there
was
a
collision
with
some
other
iitf
activities
among
others.
A
It's
a
big
group,
though,
so
it
was
somewhat
challenging
when
we
went
by
the
doodle
poll
results
and
it
could
have
perhaps
improved
this
by
being
more
aware
of,
what's
going
on
elsewhere
in
the
iitf,
but
yeah,
maybe
maybe
in
the
end
we
do
have
reasonable
number
of
people
now.
Quite
many
actually
joined
this
one.
So
that's
great,
thank
you
and
it's
also
not
not
the
greatest
time
for
some
parts
of
the
world
so
abilities
for
that.
A
As
well
brief
talk
about
the
rational
I
think
we
all
understand
so
no
need
to
go
into
the
details
of
this,
but
obviously
saving
the
planet
is
is
one
of
the
biggest
things
that
we
we
could
be
worried
about
or
working
towards
for,
and
you
know
it
it.
Just
it's
hugely
important,
of
course,
and
the
one
thing
that
I
would
like
to
highlight
is
also
that
that
it's
not
just
we're
not
just
doing
this
for
the
kid
of
the
humankind
and
good
of
the
planet,
but
this
is
also
very
much.
A
The
reality
in
business
world
for
many
of
the
private
Enterprises
have
very
hard
requirements
on
how
much
energy,
for
instance,
they
can
use,
and
what
type
so
there's
a
lot
of
reasons
like
everything
is
pointing
to
the
same
direction
that
we
have
to
worry
about
this
and
make
sure
that
we
do
our
best
and,
of
course
the
internet
is
an
incredibly
powerful
tool
and
it
can
both
help
the
societies
and
also
damage
them.
A
So
it
can
help,
for
instance,
by
as
an
example,
you
know,
organizing
meetings
that
where
people
don't
have
to
travel
on
airplanes
all
over
the
place
to
to
have
a
discussion
like
like
this
one
is,
but
it
can
also
amplify
a
harmful
issues.
So,
with
the
click
of
a
button
you
can
you
can
order
physical
Goods
on
your
doorstep
and
you
know-
maybe
that's
not
always
great
and
of
course,
the
internet
itself
uses
energy
or
the
whole
system.
A
We're
thinking
about
this
in
an
end-to-end
fashion
consumes
energy
consumes
raw
materials,
and
you
know
leaves
some
pollution
and
the
question
is
then:
can
we
do
something
about
it?
Basically,
we
are
researchers
or
technologists
or
networking
people.
A
A
Practicalities,
of
course
the
goal
is
to
learn
from
each
other.
So
let's
read
and
listen
and
comment-
and
this
is
not
a
bunch
of
Talking
Heads
I
hope,
but
it's
actual
discussion.
So
please
comment
whenever
it's
when
I
know
you
have
a
something
that's
just
the
same,
and
we
have
this
convention
that
we'll
form
a
queue
on
on
D,
because
there's
quite
many
people,
so
maybe
it's
useful
to
to
have
a
queue.
So
that
not
everybody
is
talking
at
the
same
time.
A
A
We
do
have
a
bit
of
a
like.
You
know
set
of
short
percentages
at
the
beginning.
Always
so
you
may
consider
if
you
want
to
have
your
question
in
the
middle
of
of
the
presentation.
That's
also
totally
fine
or
you
want
to
say
very
big,
maybe
a
bigger
topic
for
for
the
discussion.
Parts.
What's
sort
of
up
to
you,
everybody
should
contribute.
This
is
not
the
IAB
or
anybody
in
the
program,
quality
or
or
anyone
to
dictate
what
the
result
is.
It's
all
up
to
us.
A
So
whatever
comes
out
of
this,
what
are
the
conclusions?
It's
really
up
to
each
and
every
one
of
us
so
please
contribute
I
did
mention
already
that
the
position
papers
and
recordings
will
be
published.
A
We'll
also
produce
a
report,
as
it's
traditionally
done
from
this
IAD
workshops,
so
that
would
typically
contain
things
like
you
know,
recorder
of
the
discussions,
but
also
things
like
recommendations
or
observations,
and
these
are
hopefully
permanent
things,
so
they
will
go
into
an
RFC,
obviously,
and
everything
all
the
materials
will
be
stored
in
the
IEP
web
page
for
for
this
Workshop.
A
So
that's
the
that's
basically
the
setup,
there's
any
questions
on
on
our
comments
on
on
this
organizational
part,
and
this
would
be
a
good
time
to
ask.
C
A
Or
that's
at
least
the
scope
and
then
we'll
actually
get
to
that
in
a
in
in
one
minute,
but
but
it
depends
a
little
bit
on
what
people
are
sort
of
more
most
interested
in
it's
it's
not
that
the
scope
restricts
you,
but
you
might
be.
If
you're
a
protocol
engineer,
you
might
be
more
interested
in
political
details.
A
Okay,
so
then
I'm
gonna
move
to
the
second
presentation,
which
is
about
the
big
picture.
So
this
is
a
little
bit
about
what
John
you
were
asking
about.
So
what's
the
scope,
what
are
the
issues?
A
What
are
the
relationships?
What
is
this
only
about
technology
or
or
costs
and
I
I,
don't
know
if
this
is
sort
of
a
useful
thing.
A
I
think
it
illustrates
nicely
what
how
broad
this
topic
is,
and
so
this
isn't
like
an
official
categorization
of
everything
related
to
improving
Energy,
Efficiency
or
environmental
impacts
in
the
whole
world,
or
or
even
in
ICT
or
or
the
internet.
But
it's
it's
a
simply
a
drawing
of
you.
What
your
position
papers
fell
on
like
different
categories
and
it's
obviously
a
snapshot
of
a
particular
Community,
that's
interested
in
particular
things.
So,
for
instance,
you
see
that
the
protocol
aspects
are
fairly
by
the
focused
on
here.
A
So
there's
quite
a
lot
of
contributions
on
that
angle,
but
there's
also
another
big
item
around
this
understanding
and
measurements
and
yeah
sort
of
perhaps
from
a
slightly
more
academic
perspective
that
we
want
to
understand.
What's
going
on
how
much
you
know
this
or
that
is
consumed,
or
how
much
impact
do
we
cause
in
terms
of
raw
materials
or
or
other
other
types
of
things?
A
That's
also
important
important,
but
there's
also
tons
of
other
things.
That's
some
people
have
looked
at
implementation
aspects.
Some
people
have
looked
at
user
Behavior,
there's
the
social
societal
issues,
this
benefits
to
other
fields,
there's
even
discussion
of
some
of
or
in
some
papers
there
was
some
discussion
of
them
actors
who,
who
should
do
what
and
who
are
involved
and
so
on.
So
it's
it's
pretty
broad.
A
So
the
main
main
point
that
I'm
trying
to
make
here
is
that
this
is
this
is
bigger
than
our
individual
and
also
if
I'm
I'm
used
to
doing
a
particular
thing.
Let's
say
I'm
I'm
working
with
clean
energy,
then
you
know
I'm,
obviously
very
interested
in
making
sure
that
everybody
transits
to
or
switches
to
use
of
clean
energy,
but
that's
obviously
not
the
only
thing
that
needs
to
be
done.
A
This
multiple
aspects-
and
maybe
that's
the
thing
that
I
want
you
to
remember
so
we're
trying
to
take
an
end-to-end
view
across
devices
networks,
data,
centers
and
applications.
A
A
We
should
worry
about
greenhouse
gases,
obviously,
but
also
other
issues
might
might
be
quite
relevant.
We
should
consider
both
first
and
second
order
impacts,
so
in
a
first
order.
Impact
is,
is
that
we,
for
instance,
produce
some
pollution
or
CO2
from
running
our
systems,
but
then
there's
the
second
order
that
like
well.
We
can,
for
instance,
help
the
society
to
do
something.
You
know
more
efficient
manner
avoid
flying,
for
instance,
for
some
sense.
A
Of
course,
those
impacts
can
be
both
positive
and
negative.
So
we
have
to
remember
that
improvements
are
obviously
interesting
for
this
Workshop,
but
not
only
that,
it's
not
just
about
fixing
stuff,
it's
about
also
about
understanding
stuff.
So
if
we
at
least
understand
that
here
is
an
issue,
then
then
that's
you
know
we're
one
step
further
and
it's
also
important
to
understand
the
trade-offs
and
costs
of
you
know
whatever
we
would
change.
A
That's
oliveiro
and
others
commented
that
on
on
their
position,
paper
which
I
recommend
reading
and
we
can't
include
social
business
and
other
non-technical
aspects
and
or
even
regulation
of
governments.
So
this
is
all
sort
of
linked
solutions.
They
can
come
from
many
many
different
angles.
It
can
be
about
implementations,
protocols
and
standards,
which
are
you
know
for
many
of
us.
It's
the
main
thing.
It
can
also
be
about
clean
energy.
It
can
be
changes
in
business
practices.
You
know.
A
We
could
have
better
awareness,
better
measurements
and
better
transparency
in
the
networks
that
hey
this
is.
This
is
what's
happening
because
of
your
request,
or
we
could
simply
be
building
better
tools
for
somebody
else
to
do
a
conferencing,
for
instance,.
A
So
that's
it
from
my
side.
If
there's
any
quick
comments
on
this,
we
can
take
that.
Otherwise
we
can
proceed
to
tour,
less
understand.
D
Okay
small
fight
with
the
unmute
button,
a
classic
all
right.
Can
you
see
the
slides
hello?
Yes,
okay,
yeah
all
right,
so
let's
look
into
the
review
mirror.
So
this
is
a
position
paper
on
what
we
have
done
around
energy
in
the
ietf,
and
this
intro
slide
shows
who's
been
working
on
that
and
where
you
can
find
and
discuss
it
as
well.
D
So
why
did
we
start
this
work
now
so
last
year,
IP
RFC
7091
turned
40
years
old
and
we
didn't
have
a
party.
Of
course
there
was
covet
so
maybe
time
for
an
early
midlife
crisis.
D
So
there's
a
lot
of
positive
things.
Obviously
we're
aware
of
what
oil
technology
has
done
right
and
that
I
think
that
became
particularly
obvious
in
in
during
Corona
when
nobody
could
physically
move,
and
it
was
really
the
internet,
another
TCP
networks
that
enabled
you
know
for
society
to
really
continue
to
to
operate
much
better
than
it
would
have
been
possible
30
years
ago
or
well.
D
1918
I
think
it
was
so,
but,
on
the
other
hand,
I
think
we're
also
aware
of
a
lot
of
the
counterpoints
of
the
overall
energy
consumption
and
its
impact
in
the
I.T
industry
and
then,
of
course,
also
a
lot
of
societal,
cultural
and
political
impacts
of
the
internet
itself.
D
So
before
looking
forward,
it
seemed
good
to
start
analyzing
what
we
have
done,
and
the
hope,
of
course,
is
that
it
enables
more
contributors
for
future
work
to
understand
where
we
are,
what
we
have
done
find
in
network
gaps,
but
of
course,
as
people
with
cool
slogans
have
told
us
for
hundreds
of
years.
D
Just
because
we
know
what
the
past
is
doesn't
mean
that
you
know
it.
We
know
exactly
what
to
do
next.
D
So
what's
the
scope
and
how
do
we
go
go
along
right
so,
especially
amongst
my
colleagues,
some
of
them,
like
Hesham,
also
were
doing
a
great
job
in
collecting
a
lot
of
cool
information
about
the
whole
energy
consumption
in
the
I.T
industry
operations,
life
cycle,
the
cost
of
the
network
infrastructure
versus
the
client,
Data,
Center,
Mobile
use
cases.
So
it's
it's
really
overwhelming.
D
If
you
think
about
the
impact
of
energy
in
it
and
society
around
it,
so
of
course
it
became
a
lot
easier
when
looking
at
this
document
by
simply
saying
well,
we'll
have
to
somehow
prove
what
the
ietf
had
to
do
with
it,
by
relating
it
to
the
work
that
the
iitf
has
done,
which
is
finished,
rfcs
or
also
attempted,
and
as
of
yet
abandoned
drafts,
and
then
basically
try
to
structure
the
work
around
that.
That
makes
the
document
a
little
bit
hard
to
read
with
all
the
references
in
it,
but
obviously
for
people.
D
You
know
that
we're
targeting,
like
the
ones
hopefully
doing
new
work
with
it
I
think
that's
are
the
best
that
we
can
do,
and
it
turns
out
that
most
of
the
impact
that
our
work
has
done
is
is
really
incidental
right,
because
energy
often
became
a
relevant
metric
of
considerations
much
later
than
the
work
itself
and
intentional
energy
related
work
only
recently
happened
in
the
last,
maybe
20
years.
Right
and
one
of
the
other
points
was
you
know
to
recognize
that
we
need
to
expand
from
energy
to
sustainability
by
using
carbon-based
metrics.
D
Whenever
the
the
use
case
and
workflows
of
the
technology
really
allowed
to
take
that
into
consideration,
because
that
pretty
much
means
that
the
same
energy
consumption
is
less
problematic
if
it
can
use
renewable
energy
than
if
it
would
have
to
use,
you
know:
Carbon,
burning
energy,
all
right,
so
here
is
just
a
single
page
overview
of
what
the
document
entails.
D
It's
40
pages,
so
you
know
I
hope
many
of
you
will
have
read
it
or
will
read
it
after
this,
these
slides
it
starts
with
really
the
high
level
tenets
of
the
internet
and
other
TCP
networks
of
the
its
architecture,
its
networks,
right
and
that
is
in
general,
energy
saving
through
scale
right,
and
there
are
a
lot
of
technologies
that
produce
scale
through
convergence
through
Global
networking
through
Federation.
You
know,
the
internet
itself
is
the
biggest
Federated
Network.
There
is
encryption
and
freedom
to
innovate
right.
D
All
these
things
brought
more
applications
more
users,
more
workflows
into
the
same
networks,
allow
them
to
scale
reduce
cost
and
therefore
make
the
solutions
even
more
competitive
and
attractive
than
you
know,
their
prior
Alternatives.
If
they
then
existed
and
I'm
going
into
especially
details
of
applications,
where
mostly
aware
of
like
the
telecollaboration,
what
we've
seen
you
know,
a
40-year
history
from
email
through
telephony
like
say,
and
then
lately
RTC
web,
which
is
at
the
core
of
all
the
video
conferencing
that
you
know
has
I'd
say,
has
saved
the
planet's
Communications
during
Corona
times
right.
D
But
of
course,
the
more
broader
topic
of
digitization
replacement
of
workflows
without
it
with
those
that
use
the
internet
or
other
TCP
networks,
is
also
captured.
D
There
then
exactly
that
expansion
from
energy
saving
as
the
core
benefit
to
this,
to
enabling
the
use
of
sustainable
energy
and
computation
with
that,
for
example,
so
I
I
give
some
examples
of
how
the
internet
and
ITF
Technologies
benefit
there
and
then,
ultimately,
we
get
into
the
area
where
energy
was
exactly
one
of
the
core
targets
of
improving
the
ITF
protocol
landscape,
and
that
is
what
we
call
the
low
power
and
lossy
constrained
networks,
and
we
have
a
humongous
set
of
working
groups
that
have
been
working
on
that
and
it
all
comes
from
devices
with
very
little
ability
to
have
energy,
whether
it's
battery
based
or
self-production
of
the
energy.
D
So
there
is
a
really
good
great
amount
of
work,
which
I
think
will
also
proliferate
into
faster
and
more
energy
available
networks
in
the
future,
just
because
they're
all
more
efficient
than
their
prior
counterparts.
D
Then.
Finally,
a
lot
of
Technology
details
with
sample
technologies
that
save
energy
sleepy
nodes,
multicast
discussions
around
the
use
of
tcpip
Technologies
in
the
energy
networks
themselves,
the
smart
grid,
the
central
phaser
networks,
which
enable
really
to
avoid
brownouts
on
the
planet
much
better
than
in
the
past
working
groups
for
measurements
that
we
had
Eman
and
metrics
and
benchmarking,
and
then
the
power
awareness
of
these
metrics,
which
you
know
hasn't
gone
very
far
today
via
the
pennant
effort
and
ultimately,
of
course,
what's
happening
today-
is
through
sdn
mechanisms.
D
So
that's
kind
of
the
The
Rundown
I
haven't
counted
I,
think
it's
in
excess
of
100
references
to
to
rfcs
and
drafts
that
are
in
the
document.
So
there
is
a
humongous
amount
of
work
that
we've
done,
at
least
when
you
start
thinking
about
how
the
energy
related
impact
of
ietf
work.
Specific
documents
is
not
only
the
networks
that
were
produced
with
them
all
right,
so
reviewers
ask
me:
okay,
so
rear
view
mirror
is
nice.
But
what
really
is
next
and
I
was
trying
to
figure
out
here.
D
Three
points
right
so,
first
of
all,
I
think
they're
significant,
but
not
easy,
optimization
and
enhancement
at
the
network.
Layer,
I
think
they're.
A
lot
more.
You
know
workflows
as
we've
seen
with
RTC
web
latest
and
others
that
I
think.
If
we're
engaged
in
these
workflows
using
the
network,
they
can
become
better.
D
But
ultimately,
when
we
look
at
the
inflection
points,
the
the
really
big
changes
in
in
our
history
of
the
internet
I
think
we
have
the
internet
itself.
Mobility
data
centers,
as
maybe
the
top
three
and
I
think
the
next
and
already
starting
big
inflection
point-
is
renewable
energy
because
it
will
break
barriers
for
energy
consumptions,
maybe
only
decades
out,
but
we're
really
seeing
the
Starting
In
Pockets
I
think
it's
going
to
go
through
all
forms
of
energy
consumption
and
many
of
those
will
need
it.
D
Components
for
that
and
I've
been
giving
a
few
examples
like
the
time
and
place
shift,
but
ultimately,
when
they're
is
a
surplus
of
energy,
what
are
the
next
cool?
You
know
it
internet
based
use
cases
workflows
that
those
would
enable
all
right
and
that's
it.
Thank
you.
A
We
see
your
screen,
yes,
thank
you.
Let's.
G
E
I'm
a
little
confused
here
but
I
have
a
lot
of
stuff
popped
up
on
my
screen.
So
are
you
seeing
the
notes?
Are
you
seeing
the
main
screen.
E
E
Okay,
but
anyway,
suffice
to
say
I'm,
Eve,
Schuler,
I'm
from
Intel
and
I'm
I
have
done
this
work
with
my
colleagues
from
Oxford
and
Ori
and
Yale
and
Intel,
and
we
co-authored
the
position
paper
perspective
on
carbon
aware
networking
and
the
request
for
this
lightning
talk,
however,
was
really
to
focus
on
carbon
versus
energy,
and
that's
really
because
if
you
look
at
sustainability,
research
within
ICT
Community,
the
information
communication,
technology,
Community
or
sector,
it's
traditionally
focused
on
Energy
Efficiency
and
the
goal
to
consume
this
energy.
E
The
effective.
The
most
effective
action,
of
course,
is
at
the
intersection
of
this
Venn
diagram
and,
more
typically
or
I.
Would
say,
paradoxically,
it's
often
the
case,
one
pillar
is
achieved,
but
not
necessarily
the
others
so
for,
for
example,
zero
emissions
does
not
necessarily
equal
energy
efficient
case
in
point
being
some
of
what's
going
on
with
renewably
powered
Bitcoin
mining
and
cryptocurrency
data
centers.
E
Nor
does
Energy
Efficiency
equal
zero
emissions.
A
high
Energy
Efficiency
Factory
may
be
located
in
a
geography
where
the
power
is
generated
from
fossil
fuel
and
therefore
has
carbon
high
carbon
intensity,
and
so
it
really
again
I'm
struggling
a
little
bit
with
my
screens
here,
but
so
to
really
reduce
the
growing
electricity
usage
of
the
ICT
sector
and
in
turn,
the
carbon
footprint
of
data
centers.
E
Some
of
the
hyperscalers
have
developed
and
deployed
what
they
are
calling
either
carbon
intelligent
or
carbon
aware
Computing,
and
that's
really
focusing
on
maximizing
the
usage
of
renewable
and
clean
energy,
and
they
time
shift
their
compute
orchestrated
workloads
to
align
with
when
electricity
has
the
lowest
carbon
intensity
and
is
the
cleanest.
And
this
is
a
really
nice
symbiotic
relationship
in
that
Renewables
help.
E
E
But
it,
more
importantly,
leads
us
to
ask:
why
not
employ
carbon
awareness
everywhere
and
throughout
the
entire
ecosystem,
ICT
ecosystem,
and
not
just
in
hybrid
hyperscaler
data
centers,
but
throughout
the
edge
to
Cloud
Continuum
throughout,
as
we
begin
to
roll
out
next
Generation
infrastructure,
particularly
in
places
where
it's
possible
to
more
approximately
co-locate
with
Renewables?
E
Let's
not
just
focus
on
carbon
aware
compute,
but
other
facets
of
systems
like
storage
and
networking,
and
begin
to
adopt
carbon
intestine
intensity
as
one
of
these
quality
of
service
metrics,
and
we,
of
course
want
to
look
at
not
just
what
can
we
do
with
Hardware,
but
all
the
way
through
software
components
vertically
up
and
down
the
stack
and
in
one
software
that's
resident
in
a
particular
platform,
but
also
end-to-end
horizontally,
and
we
are
going
to
need
a
lot
entry
points
in
the
architecture
for
apis
that
allow
us
to
gain
some
insights
from
carbon
awareness.
E
I've
included
a
couple
graphs
on
the
right
and
they're
blown
up
in
the
background
of
my
slide,
so
you
can
peruse
the
details,
but
here
the
main
point
was
that
the
graph
on
the
top
emphasizes
pictorially
sort
of
the
growth,
and
we
can
argue
about
how
much
growth
is
going
to
happen.
But
we
do
believe
that
growth
is
going
to
continue.
E
The
the
orange
in
that
is
the
data
center
growth.
But
you
can
see,
there's
a
lot
that
lies
outside
of
the
data
centers
and
when
you
study
the
details,
you
can
see
that
networking
is
definitely
on
par
with
data
centers
in
terms
of
its
carbon
footprint.
E
So
there's
some
self-reflection
we
should
be
doing,
but
the
the
bottom
graph
is
actually
from
California's
estimation
of
it's
not
estimation,
it's
the
measurement
of
it's,
how
much
curtailment
it
has
to
do,
because
it
had
excess,
renewable
energy,
and
you
can
see
the
variability
and
growth
of
that
as
well.
E
E
I'm
going
to
sort
of
quickly
go
through
this
for
carbon,
aware
routing
we'd,
really
like
our
current
routing
protocols
to
be
able
to
select
we'd
like
them
to
evolve,
to
be
able
to
select
more
carbon
efficient
paths,
possibly
by
considering
carbon
intensity
as
an
additional
qos
metric,
but
also
by
acknowledging
that
we
have
time
variant,
links
some
that
might
be
powered
by
renewable
energy
and
thus
might
predictably
come
be
coming
and
going
inspired
by
some
of
the
delay.
E
Inspired
I
think
by
some
of
the
work
going
on
in
the
debt
net
or
deterministic
networking
community
carbonware
traffic
engineering
could
be
called
upon
to
guarantee
that
flows
stay
within
carbon
consumption
budgets,
possibly
employing
hot
by
hop
techniques
and
possibly
also
reserving
clean
energy
resources
along
the
way.
And,
of
course,
foundational
to
everything
is
that
carbon
Telemetry
needs
to
exist
for
all
this
work,
and
it
requires
Network
element
observability,
but
also
an
awareness
and
reduction
of
its
own
impact.
E
And
again,
due
to
time
here,
we'll
we'll
cue
up,
you
know
what
the
challenges
are
here
for
discussion
in
session.
E
Four
we're
going
to
have
another
lightning
talk
to
talk
about
the
the
next
steps
for
carbon,
aware
networking,
but
suffice
to
say
here
that
some
of
the
challenges
are
really
to
make
real
time,
tracking
and
reporting
for
electricity
consumption,
something
that's
accessible
additionally,
tracking
and
reporting
for
finer
grain,
electricity,
carbon
intensity
and
those
things
together,
map
to
the
the
network
paths
that
we
use,
because,
ultimately,
if
we
want
to
understand
the
carbon
efficiency
or
the
environmental
impact
of
applications
and
services,
we
need
to
understand
the
networking
paths
that
they're
all
using,
especially
as
they
become
more
distributed
in
time,
which
they
have
and
and
obviously
once
you
have
all
that
information.
E
E
I,
this
is
just
to
point
you
at
a
couple
other
readings.
If
you
want
to
look
at
further
details,
but
then
there's
some
terrific
papers
throughout
the
workshop
having
to
do
with
carbon
in
general,
whether
that's
Bruce
nordman's
paper
on
pricing
of,
what's
going
on,
you
know
behind
the
scenes,
the
sort
of
business
elements
there's
additional
there.
There's
another
paper
on
carbon,
aware
networking,
the
Chris,
Adams
and
and
his
co-authors
wrote
as
well
as
interesting
papers
on
metrics.
E
E
C
A
Go
I
think
yeah
good.
Thank
you
that
was
very
interesting.
I
think
I
will
at
least
have
some
questions,
but
some
other
questions
on
the
chat,
but
maybe
we'll
go
to
Vesna.
First,
all
right,
you
online,
and
can
you
share?
You
also
have
a
different
and
very
interesting
angle.
There
we
go.
Thank
you.
B
And
that's
not
I
work
as
a
community
builder
for
ripe
NCC.
However,
today
I'm
speaking
as
an
activist
and
as
an
intersectional
feminist
and
as
a
parent
of
activists,
so
I
submitted
the
paper
that
have
been
seen
very
as
very
political,
so
I've
been
asked
to
talk
more
about
the
a
societal
aspects
of
where
does
the
environmental
sustainability
fit
not
only
from
the
technical
perspective,
and
so
I
wish
to
propose
that
we
don't
limit
ourselves
to
only
talking
about
environmental
sustainability,
but
work
towards
climate
Justice.
B
So
I'd
like
to
on
this
one
slide
propose
like
three
problems
and
three
possible
Solutions
in
a
broadest
possible
sense.
B
B
But
now
we
can
actually
also
call
a
digital
colonialism
in
which
we
are
kind
of
repeating
the
history
of
the
actual
economic
colonialism
in
the
digital
sphere,
where
we
copied
the
exploitations
economic
exploitation
and
nature
exploitation
that
are
predominantly
kind
of
externalize
towards
the
global
South
and
cause
most
harms
to
the
communities
that
are
actually
removed
from
all
these
decision-making
processes
and
from
the
benefits
that
digitalization
brings
to
the
people
in
the
global
North
North
and
another
part
of
that
is
the
policies
of
extractivism,
so
just
taking
as
much
energy
resources,
raw
materials
water
in
order
to
produce
these
digital
devices
and
then
later
kind
of
throwing
them
away
and
again,
predominantly
in
the
global
south
or
the
in
an
in
unequal
way.
B
The
way
how
to
deal
with
that
would
be
to
consider
ourselves
as
one
humanity
and
to
show
solidarity
with
the
the
parts
of
the
world
and
the
communities
that
have
been
already
impacted
disproportionately
with
these
developments
and
to
at
least
minimize
the
harms
and
with
the
goal
of
stopping
the
harms
to
those
communities
and
and
those
natural
environments.
B
The
second
problem
is
our
belief
in
the
endless
growth
on
a
a
limited
Planet
which,
as
engineers
and
scientists,
we
should
understand
that
it
is
just
impossible,
and
so
that
is
seen
in
consumerism
or
chasing
the
luxury
goods
or
in
the
more
like
technical
optimism
way
as
going
for
Innovation
trying
to
make
things
to
be
faster
and
larger
and
reaching
more
and
more
of
the
consumers,
let's
say,
and
also
in
the
short
term
thinking
and
aiming
for
convenience
and
so
to
to
combat
that.
B
We
need
to
change
the
way
of
thinking
and
to
implement
the
the
thinking
of
sufficiency
or
even
modesty
and
frugality
in
our
technical
design,
considering
limiting
extractivism,
decreasing
the
growth
and
focusing
on
the
common
resources
and
common
goods,
or
in
more
like
down
to
Earth
Tech
on
the
free
and
open
source
software
and
open
hardware
and
Open
Standards.
That
ITF
is
famous
for
and
finally
addiction
to
fossil
fuels
and
all
the
other
over
consumed
materials
like
water
and
land
and
minerals.
B
We
have
to
just
stop
doing
that,
so
our
previous
speakers
already
talked
about
The
Greening,
the
energy
sources,
but
we
also
have
to
decrease
the
use
of
energy,
even
if
it
is
renewable
energy
because
it
is
needed
for
the
more
life
affirming
needs
of
the
people
rather
than
for
the
data
centers,
and
so
in
the
tech
field.
We
should
be
focusing
on
repairability,
circular
economy
and
durable
Tech.
B
My
next
slide
is
just
translating
these
suggestions
for
the
follow-up
sessions
on
this
workshops.
So
in
the
second
session
there
were
a
lot
of
topics
covered
about
the
energy,
but
I
would
like
to
suggest
also
considering
the
water,
minerals,
consumption
and
focus
Less
on
the
measurements
and
more
on
the
actual
activities.
B
We
want
to
decrease
everything
by,
let's
say
10
percent
per
year,
and
the
other
advice
would
be
to
consider
that
the
future
is
going
to
contain
more
protests,
more
Wars,
more
refugees,
more
hurricanes
and
other
national
natural
disasters
that
we
should
adjust
the
tech
to
those
emergency
situations
rather
than
to
the
luxury
usages
that
we
have
been
accustomed
to
and
to
be
aware
of
the
efficiency,
paradoxes
and
the
power
structures
and
the
power
differentials
to
consider
who
is
not
in
the
room
and
how
can
we
consider
their
needs?
B
And
finally,
as
the
next
steps,
please,
let's
not
limit
ourselves
to
just
us
who
are
in
the
room
but
consider
working
together
with
both
the
end
users.
There
is
already
an
RFC
about
that
how
the
internet
is
for
the
end
users,
the
research
groups
like
Gaia
and
activists,
and
also
the
existing
climate
Justice
organizations.
B
And
there
won't
be
any
internet
on
the
burning
Planet.
So
that's
why
I'm
speaking
here
as
an
activist,
thank
you.
A
So
now
we
have
about
15
minutes
for
discussion
if
you
want
on
the
Queue
just
put
plus
Q
in
the
in
the
chat
and
actually
I
did
that
as
an
example,
and
also
because
I
wanted
to
ask
a
question
so
I'll
shoot
my
question.
First
Eve
just
inspired
by
your
presentation,
which
was
very
interesting.
A
So
so
you
talked
about
this
carbon
awareness
and
is
that
what
the
the
approaches
that
would
that
be
mostly
interdomain
or
interdomain,
because
I
presume
that
the
challenges
between
those
two
would
be
very
different,
like
you
can
in
interdoma,
and
you
can
trust
mostly
what
the
devices
say
about
interdometer
might
be
much
more
difficult
right.
E
E
I
I
don't
know
where,
where
it
will
be,
but
I
mean
you've
targeted,
something
that
we've
been
thinking
about,
which
is
within
an
administrative
domain.
E
You
might
have
the
granularity
of
detail
about
what
not
only
the
devices
along
a
path,
all
those
Network
elements
along
the
path
might
be
consuming
and
where
they
reside
and
all
of
that
interesting
Rich
information,
but
once
you
leave
the
bound
that
boundary
or
that
scope,
you
are
at
the
mercy
of
you,
know
other
elements
and
and
not
only
that,
but
there's
so
many
elements
that
are
hidden
from
view
in
our
Network
paths
that
also
consume
energy
that
have
these
secondary
effects.
E
So
is
it
sufficient
to
just
measure
the
the
consumption
of
energy
along
a
route?
There's
like
this
whole
Supporting
Cast
that
we
don't
quite
know
how
to
comprehend?
Just
yet,
and
so
we
are
going
to
have
to
partner
with
other
as's
that
we
Transit
and
so
I.
We
don't
have
a
solution
as
yet
I
mean
I.
Think
that's
where
the
research
is,
what
should
be
exposed
should
it
be?
E
Where
should
it
be
exposed
should
be
exposed
at
the
boundaries
so
that
for
hop
by
hop
technology,
we
can
kind
of
accumulate
ranges
of
expectation
around
energy
usage
and
carbon
intensities,
and
things
like
that,
so
it's
a
big
unknown,
and
so
we
welcome
people's
thoughts
either
off
the
cuff
or
if
they
have
experience
in
things
that
are
similar
to
this,
for
example,.
G
Hello,
so
this
was
really
interesting
and
and
quite
diverse,
so
I
actually
sort
of
felt
that
vesnar
sort
of
summarized
a
bunch
of
points
that
that
I've
sort
of
been
been
bouncing
around
in
my
head
right,
because
you
know,
obviously
we
can
try
and
build
an
Internet
that
uses
like
half
the
power
right.
We
rip
all
the
equipment
out
and
we
put
new
equipment
and
then
you
know,
then
we
were
great.
G
We
were,
you
know
only
using
half
the
power,
but
but
in
terms
of
waste
and
CO2
balance,
we've
probably
done
pretty
poorly
because
we,
you
know
just
doubled
the
embodied
carbon
of
the
internet
and
so
I
think
we
got
to
get
to
the
point
that
that's
now
is
lining
out
and
she
did
it
in
a
very
activist
way.
But
I
really
feel
sympathetic
to
the
concept.
We
need
a
sort
of
not
look
at
the
internet
and
its
capacity
is
unlimited
right.
G
We've
always
sort
of
been
banking
on
the
internet
getting
better
over
time
and
and
it
does
that
by
using
more
and
more
resources
and
I
think
we
gotta
sort
of
slow
that
down
or
ideally
reverse
it.
That
means
we
need
to
sort
of
figure
out
what
is
worth
sending
when
and
what
is
worth
putting
on
internet
and
what
isn't-
and
this
isn't,
unfortunately,
just
us
in
the
ITF
right-
it's
we
don't
generate
the
data.
G
The
data
gets
generated
by
various
services,
some
platforms,
and
what
have
you
but
I
think
this
sort
of
notion
that
you
know
we
got
to
be
more
frugal
with,
with
the
capacity
might
lead
us
down
a
path
that
sort
of
hopefully
gets
us
to
a
burning,
Planet
slower.
Thank
you.
A
Thanks
after
you,
it
will
be
sures
and
pernila.
A
H
Not
hearing
you
sorry
usual
error.
One
thing
that
I
always
wonder
is
that
I
don't
know
how
to
put
this.
But
what
happens
if
all
energy
is
from
your
renewable
sources?
Is
that
the
end
of
the
work
and
because
it
seems
to
be
that
you
know
a
lot
of
work
is
done
on
kind
of
claiming
that
if
we
are
using
renewable
energy,
then
all
is
fine.
H
So
we're
talking
about
give
or
take
500
percent
of
more
renewable
energy
needed
to
be
built
for
this
one
single
player,
so
renewable
doesn't
mean
carbon
free
and
that's
the
point
that
you
know
that,
in
my
opinion,
the
answer
that,
with
your
renewable
energy,
we're
solving
the
problem,
then
that's
the
kind
of
the
short-term
solution
for
the
next
few
years.
And
after
that
you
know,
what's
the
next
solution,
so
kind
of
I
was
pretty
much
in
line
with
the
West
numbers.
Western
I
was
saying
and
there's
many
many
topics.
H
I
Thank
you.
Thank
you
like,
so
that
was
really
fascinating
presentation.
So
one
question
I
had
is
like
you're
talking
about
how
by
hop
stuff,
do
you
know
of
anything
that,
like
you
know
how
these
like
hops,
are
abortioned
their
share
of
like
carbon
emissions
like?
Is
there
something
we
can
actually
do
there
right,
because
it
seems
like
a
very
difficult
Problem,
Like
We
Never
managed
to
do
this
like
in
a
multi-domain
basis
or
multi-administrative
domain
basis.
So
is
there
any
thoughts
like
how
you
would
go
about
it?
Okay,.
E
I
I
was
drawing
analogies
to
sort
of
what
happens
in
the
measurement
Community,
as
you
Transit
paths
and
you
collect
information
measurement
information,
and
so
you
know
I,
don't
know
whether
that
those
might
be
the
same
kinds
of
technologies
that
we
we
would
use
or
protocols
that
do
that.
E
But
you're
right,
we
haven't
traditionally
looked
beyond
specific
kinds
of
attributes
about
our
Network
performance,
like
packet,
loss
and
latency,
and
the
variance
in
the
latency
otherwise
known
as
gender,
and
it
begs
the
question:
can
we
even
rely
on
external
measurement
from
other
places,
to
tell
us
about
the
energy
consumed
or
the
carbon
intensity,
which
are
things
that
might
fall
outside
the
ownership
of
those
who
are
operating
the
networks?
E
And
so
we
need
to
have
some
kind
of
trust
relationship
there
I,
don't
so
I
can't
say
that
you
know
which
technologies
have
really
scaled
in
order
to
do
hop
by
hop
accumulation.
I
know
that
there
have
been
and
Carlos
please
pipe
up.
I
know
that
you
did
this
canvassing
I
sort
of
view
RSVP
as
one
of
these
protocols
that
that
did
some
of
that
there
are
other
things
out
there.
E
I
was
thinking
more
in
the
small
scale
looking
to
the
debt
network,
but
there's
lots
of
things
coming
out
of
the
t's
working
group
and
elsewhere
in
the
measurement
side
of
the
ITF.
That
might
be
places
to
glean
for
insights
there.
E
F
Hi
all
sorry
I
couldn't
join
you
from
start
I'm,
Penny,
Lima
colleague
of
ijare
and
I
work
as
a
principal
researcher
for
ICT
sustainability
impact
and
I'm
also
co-chairing
the
standardization
work
in
I2
on
the
development
of
assessment
standards,
so
I
just
wanted
to
to
first
off
I'm
very
glad
to
be
here.
It
seems
like
a
very
interesting
event,
but
I
also
wanted
to
to
both
comment
on
what
Yuka
said
and
also
to
refer
a
bit
to
the
work
that
we
have
done
in
itu,
because
that
could
be
be
of
interest
here.
F
So
I
agree
with
when
we
so
in
in
the
frame
of
itu.
We
have
developed
a
trajectory
which
is
a
normative
and
feasible
trajectory
for
the
ICT
sector,
and
when
we
talk
about
the
ICT
sector,
we
talk
about
data
center
networks
and
user
devices
taken
from
a
life
cycle
perspective,
and
it
says
that
the
sector
should
reduce
its
emission
by
the
footprint
emissions
by
45
from
2020
to
2013..
This
has
also
been
adopted
by
the
science-based
Target
initiative.
F
So
what
what
we
have
seen
in
this
trajectory
is
that
about
80
percent
of
the
overall
life
cycle
footprint
relates
to
the
use
of
electricity,
and
so
our
course
switching
to
renewable
is
per
month
is
very
important,
but
it's
not
sufficient.
F
First,
it's
not
the
full
footprint,
it's
not
the
full
perspective
of
environmental
impacts,
and
it's
also,
as
Joker
mentioned.
It's
it.
We
we
still
need
to
produce
the
energy,
even
if
we
come
from
from
renewable
sources
and
that
we
this
will
be,
of
course,
a
constraint
in
the
availability.
F
So
there
are
many
reasons
to
to
think
broader,
but
but
this
this
trajectory
we
have
developed
to
gather
with
the
Net
Zero
standard,
which
is
aligned
with
the
such
as
the
the
race
to
cro
and
the
science-based
target
initiative
Net
series
than
that.
But
from
the
assessment
perspective
we
have
in
particular
developed
an
LCA
standard
for
for
products,
networks
and
services.
F
There
is
a
standard
for
calculating
the
footprint
of
the
entire
sector
and
what
should
be
taken
into
account
and
so
on,
and
this
is
important
because
we
see
that
many
studies
are
quite
simplistic,
let's
say,
and
they
make
assumptions
on
energy
models
which
are
just
scaled
with
data,
though
we
know
that
the
majority
of
energy
is
there
to
keep
the
systems
up
and
running
and
so
on,
and
then
we
have
also
exactly
today,
actually
agreed
or
approved,
even
a
standard
which
is
dealing
with
the
indirect
how
to
to
quantify
the
indirect
so
effects
of
ICT.
A
Thanks
and
at
least
on
my
scan
I,
don't
see
other
people
where
about
to
run
out
of
time.
So
if
people
have
quick
comments,
we
can
take
them
now.
Otherwise
we
could
finish
and
then
return
to
the
topics
on
Thursday
I
did
want
to
say
that
that,
like
this,
you
know
this
discussion
that
we
just
had
about
renewable
energy
versus
reducing
the
energy
consumption
and
so
and
I
think
that's
yeah,
I,
probably
a
trap
that
we
should
not
fall
into
too
much.
A
It's
not
a
question
of
you
know
whose
solution
can
solve
this
whole
problem.
If
we
had
a
solution
like
that,
you
know
it'd
be
much
better,
we
don't.
Actually
we
actually
have
to
use
all
of
these
tools
that
we
have
have
been
discussing
too
lesser
a
greater
extent
and
that
together
maybe
helps
us
a
little
bit
or
hopefully
gets
our
part
of
the
system,
at
least
out
of
the
critical
path.
A
But
but
it's
not
just
one
tool:
it's
yeah,
clean
energy
and
reductions
and
better
implementations,
better
standards,
taking
into
account
the
societal
issues
and
Justice
and
everything
else
so
I
think
that
that's
my
conclusion
at
least
any
other
last
words
or
or
comments.
G
A
Yeah,
indeed,
thank
you,
Lars
and
part
of
the
reason
we
have
separate
sessions
on
separate
days
is
that
we
can
have
time
between
the
systems
to
think
about
stuff,
and
you
know
reflect
and
maybe
send
email
and
also
prepare
for
the
next
next
discussions.
A
Okay,
then
thank
you.
This
has
been
a
very
interesting
session
with
many
different
angles
and
good
discussions,
so
this
clearly
work
to
be
done
in
terms
of
having
right
measurement
data.
We
had
some
discussion
of
that
on
the
chat
as
well
and
and
and
then
the
actual
improvements.
So
let's
get
to
that,
thank
you.