Minutes interim-2022-eimpactws-01: Mon 14:00
minutes-interim-2022-eimpactws-01-202212051400-00

IAB E-Impact Workshop Session 1: The big picture

5:45 am - 7:02 am  Monday, December 5, 2022 | (UTC-08:00) Pacific Time (US &
Canada)

stefano salsano
Fieke Jansen
Eve Schooler
Jari Arkko
Safiqul
John Preuß Mattsson
Julien Maisonneuve
Lars Eggert
Marisol Palmero Amador
Qin Wu
Stefano Salsano
Tirumaleswar Reddy
Eric Voit
Martin Flack
Henk Birkholz he/him
Nina Lövehagen
Shayna Robinson
Hosein Badran ISOC
Ali Rezaki
Cindy Morgan
Rick Taylor
Snezana Mitrovic
Eric Vyncke
Rob Wilton
Selome
Chiara Lombardo - CNIT  J
iankang Yao
Arif
Remy Hellstern
Beatrice Siccardi
Colin Perkins
Pascal Thubert
olivier.bonaventure
Toerless Eckert
Jan Lindblad
Vesna Manojlovic
Esther Roure Vila
Dom Robinson
Bruce Nordman
Carsten Bormann
Daniel Schien
Russ White
Romain Jacob
Mohamed Boucadair
Laurent Ciavaglia
Mike Mattera
Brendan Moran
RTP7V-3-WALLENSTEIN GARDENS
Carlos Pignataro
Chris Adams
Luis M. Contreras Telefonica
Kristin Moyer
Wim Vanderbauwhede
Per Andersson
Gonzalo Salgueiro
Dawn Nafus
Michael Welzl, University of Oslo
Suresh Krishnan
Noa Zilberman
Jukka Manner
Alex Clemm
Pernilla Bergmark Ericsson

WEBVTT

1
Jari Arkko 00:08:12.344 --> 00:08:16.874
Hello everybody can you hear me right?

2
Eric Voit 00:08:17.924 --> 00:08:18.494
Good morning.

3
Jari Arkko 00:09:31.634 --> 00:09:33.314
See, my screen yes yes.

4
Cindy Morgan 00:09:36.614 --> 00:09:37.634
Yes, we can see it.

5
Michael Welzl, University of Oslo 00:09:42.285 --> 00:09:42.975
Uh.

6
Eric Voit 00:10:51.975 --> 00:10:52.635
Want to go on mute.

7
Jari Arkko 00:11:23.895 --> 00:11:24.915
So, we'll get started.

8
Jari Arkko 00:11:39.494 --> 00:11:44.474
This thing you should hear. No, somebody was wondering about their audio.

9
Rick Taylor 00:11:47.895 --> 00:11:48.675
Working for me.

10
Ali Rezaki 00:11:50.774 --> 00:11:51.944
Yeah, it works.

11
Ali Rezaki 00:11:53.654 --> 00:11:56.804
Good morning. Good afternoon. Good.

12
Jari Arkko 00:11:56.954 --> 00:12:01.244
Daniel reboot often helps for me.

13
Jari Arkko 00:12:11.474 --> 00:12:15.284
Yeah, let's wait for 30 more seconds and then get going.

14
Jari Arkko 00:12:34.514 --> 00:12:36.224
37 people so far and so that's.

15
Brendan Moran 00:12:36.224 --> 00:12:36.764
Great.

16
Jari Arkko 00:12:38.204 --> 00:12:49.094
Yes, I think we can get started. Now, everybody's audio issues are resolved. So
welcome to the workshop for, by the.

17
Jari Arkko 00:12:49.274 --> 00:12:51.374
The environmental impacts of the Internet.

18
Jari Arkko 00:12:52.759 --> 00:13:02.474
And we're going to get started with some ground rules so we will record this
session. Cindy. Can you confirm that the recordings on, um.

19
Cindy Morgan 00:13:02.804 --> 00:13:03.824
Are being recorded.

20
Jari Arkko 00:13:04.724 --> 00:13:13.844
Yes, good. And, uh, the record is will actually be published, uh, in YouTube
later. Um, your position papers for those submitted the position paper.

21
Jari Arkko 00:13:14.864 --> 00:13:35.024
To be made public on the workshop webpage and, um, this is, of course, a
professional meeting. So we expect everybody to, uh, behave in a professional
manner. Any kind of harassment is not tolerate it. And, uh, I do want to say
that this is like, if you used to meetings and this is maybe a little bit more
diverse.

22
Jari Arkko 00:13:35.059 --> 00:13:53.324
Group of people, 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, uh, let's all learn
from each other's viewpoint. So that's the reason we actually are here.

23
Jari Arkko 00:13:55.454 --> 00:14:15.914
And, uh, today's agenda, this is the welcome session we cover to cover some
practicalities. Uh, why we are here. We can talk about, um, some big picture
issues, um, interaction elsewhere how this relates to societal issues and so
forth. You can see the agenda on screen. I hope and, um.

24
Jari Arkko 00:14:15.949 --> 00:14:24.404
Start with, uh, well, doing this, and also some of the big picture, um, issues,
and then we'll have.

25
Jari Arkko 00:14:26.114 --> 00:14:45.164
And, um, it's the same pattern for all of these sessions. We actually have 4
sessions. So this is the 1st of the 4. um, so we will have not everybody's
presenting, like, their paper. Um, most people aren't, but, uh, we will have
some, uh, hopefully exciting presentations at the beginning to.

26
Jari Arkko 00:14:45.794 --> 00:14:57.644
Get us started and then we can discuss and the discussion is actually maybe the
main goal. Perhaps we can get to some conclusions, even in some, some cases.
So, at least identify some issues.

27
Jari Arkko 00:15:00.044 --> 00:15:01.424
So far.

28
Jari Arkko 00:15:06.974 --> 00:15:22.664
Moving on, so, the upcoming sessions, uh, this week, Thursday and Friday,
there's, uh, 2 sessions 1 is on. What do we understand about the situation? So
that's about things. Like, how much is the Internet actually using, for
instance, energy or producing carbon.

29
Jari Arkko 00:15:23.745 --> 00:15:44.775
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 don't actually think that we understand
everything. So clearly, this rule for discussion, and then on the Friday. So
we'll talk about improvements of a different category.

30
Jari Arkko 00:15:46.274 --> 00:16:05.804
We'll have some, uh, talks on specific areas of technology to get us started
and we'll discuss and then on the next Monday exactly. Week from now, all the
systems 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 earliest.

31
Jari Arkko 00:16:06.284 --> 00:16:27.074
And look at specific areas, um, where are we on that then 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, um, uh, schedule
these sessions uh, if there was a coalition with some other.

32
Jari Arkko 00:16:27.080 --> 00:16:48.135
Is among others, um, it's a big group, though, so it was, um, somewhat
challenged when we went by the Doodle poll results and, uh, it could have
actually improve this by being more aware of what's going on elsewhere in the
but, uh, yeah, maybe, maybe in the end we do have reasonable number of people
now quite many.

33
Jari Arkko 00:16:48.374 --> 00:16:55.364
Join this 1, so that's great. Thank you. And it's also not not the greatest
time for some parts of the world. So, apologies for that as well.

34
Jari Arkko 00:16:57.314 --> 00:17:17.744
Um, brief talk about the rationale, I think we all understand, so no need to go
into the details of this but, uh, obviously, uh, saving the planet is, is 1 of
the biggest things that we could be, uh, worried about or working towards for,
um, and, you know, I, it just.

35
Jari Arkko 00:17:17.865 --> 00:17:38.925
It's hugely important of course. And, um, the 1 thing that I would like to
highlight is also that, that it's not just we're not just doing this for the of
the human kind, and kind of the planet. But this is also very much the reality
in business world for many of the private enterprises have very hard
requirements on how much energy for instance, they can use.

36
Jari Arkko 00:17:39.015 --> 00:18:00.135
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 we can both
help the societies. And also that makes them so it can help, for instance, by
as an example.

37
Jari Arkko 00:18:00.164 --> 00:18:21.074
Um, you know, organizing meetings where people don't have to travel on
airplanes all over the place to, to have a discussion like like this 1 is, but
it can also amplify harmful issue. So, with the click of a button, you can you
can, uh, order physical goods on your doorstep and, you know, maybe that's not
always great. Um.

38
Jari Arkko 00:18:21.464 --> 00:18:42.314
And, of course, the Internet, uh, itself uses energy or the whole system we're
thinking about this, in an end to end fashion consumers, energy consumers, raw
materials, and, um, you know, leaves and pollution. And the question is, then,
can we do something about it? Basically, we are researchers or technologists
or, um.

39
Jari Arkko 00:18:42.494 --> 00:19:02.564
Working people, um, can we improve somehow in terms of the costs of the
Internet, or provide better benefits from the Internet, the rest of the
society, or at least, could we promote research that we think is needed in
order to make improvements later or? Um, just understand things better.

40
Jari Arkko 00:19:06.765 --> 00:19:24.315
Um, practicalities, um, of course, the goal is to learn from each other. So,
let's, um, read and listen and comment and this is not a bunch of talking
heads. I hope. Um, but it's actual discussions. So please comment, uh, whenever
it's.

41
Jari Arkko 00:19:24.975 --> 00:19:45.855
Whenever you have something to just to say, and we have this convention that,
uh, we'll form a queue on on the, because there's quite many people. So, maybe
it's useful to to have a queue sort of not everybody is talking at the same
time. So, if you on the chat lines, say plus queue, then you enter the queue
and say minus queue then.

42
Jari Arkko 00:19:45.914 --> 00:19:52.304
Execute all try and track that list of people and call up when it's your time
to speak.

43
Jari Arkko 00:19:54.555 --> 00:20:12.345
We do have a bit of a like, a set of short percentages at the beginning always.
So you may consider if you're gonna have your question in the middle of of the
presentation that's also totally fine. Or you want to say, maybe maybe a bigger
topic for for the discussion parts, and sort of up to, you.

44
Jari Arkko 00:20:14.355 --> 00:20:35.235
Uh, everybody should contribute. This is not, uh, the or anybody, um, in the
program or or anyone to, um, dictate what the result is. Uh, it's all up to us.
So whatever comes out of this. What are the conclusions it's really up to each
and every 1 of us so please contribute. Um, I did mention already that the
position papers.

45
Jari Arkko 00:20:35.385 --> 00:20:56.385
And recordings will be published, we'll also produce a report, uh, as it's
traditionally done from this workshops. Um, so that would typically contain
things like, you know, recorder of the discussions. But also things like
recommendations or observations. And, um, these are hopefully.

46
Jari Arkko 00:20:56.414 --> 00:21:05.324
Permanent things, so they will go into an RFC obviously, and everything. All
the materials will be stored in the page for for this workshop.

47
Jari Arkko 00:21:08.294 --> 00:21:19.454
So, that's the, that's basically the setup, um, there's any questions on or
comments on on this, um, organizational part. Then this would be a good time to
ask.

48
John Preuß Mattsson 00:21:27.824 --> 00:21:45.614
Or we only discussing the impact, the Internet haven't negative impact, um,
Internet have on society or are we also discussing how the Internet can help
reducing the carbon footprint and other industries like like reducing travel.

49
Jari Arkko 00:21:47.444 --> 00:21:48.014
Both.

50
Jari Arkko 00:21:50.264 --> 00:21:55.784
Or that's at least the scope and then we'll actually get to that in, in in 1 a
minute but.

51
Jari Arkko 00:21:57.019 --> 00:22:08.384
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 your protocol
engineer, you might be more more interested in protocol details.

52
Jari Arkko 00:22:12.555 --> 00:22:17.595
Which is more about the costs than than helping helping other other parts of
the society.

53
Jari Arkko 00:22:20.054 --> 00:22:22.634
Any other questions shall we move forward?

54
Jari Arkko 00:22:26.504 --> 00:22:44.954
Okay, um, so then I'm going to move to the 2nd presentation, which is about,
um, the big picture. Um, so this is a little bit, uh, about what's on. You were
asking about so, what's the scope uh, what other issues or other relationships?
What is this? Only.

55
Jari Arkko 00:22:44.960 --> 00:22:48.105
But technology or or costs and, um.

56
Jari Arkko 00:22:49.274 --> 00:23:10.214
I don't know if this is sort of a useful thing. Um, I think it illustrates
nicely. What? Uh, W, how broad this topic is and so this isn't like an official
categorization of everything related to improving energy efficiency or
environmental impacts, um, in the whole world or, or even in, um, or or.

57
Jari Arkko 00:23:10.365 --> 00:23:31.485
The Internet, but it's, uh, simply a drawing of what your position papers fell
on, like different categories. And, um, 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 highly, um, focused on.

58
Jari Arkko 00:23:31.490 --> 00:23:52.605
From here, um, so there's quite a lot of contributions on that angle, but
there's also another big item around this understanding and measurements and,
um, yeah, so perhaps from a slightly more academic perspective that we want to
understand what's going on how much, you know, this or that is consumable.

59
Jari Arkko 00:23:52.784 --> 00:23:59.774
How much impact would be cross in terms of raw materials or or other other
types of things.

60
Jari Arkko 00:24:01.185 --> 00:24:22.215
It's also important important, but there's also tons of other things. There's
some people have looked at the implementation aspect. Some people have looked
at user behavior, the source society, at least use this benefits to other
fields. There's even discussion of solver, or in some papers, there was some
discussion of the actors.

61
Jari Arkko 00:24:22.274 --> 00:24:42.314
Who who should do what and who are involved and so on so it's it's pretty
broad. So the 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.

62
Jari Arkko 00:24:43.635 --> 00:24:59.205
How did you very interested in making sure that everybody transits to, uh, uh,
switches to use of clean energy? But that's obviously not the only thing that
needs to be done. That's multiple aspects. And maybe that's the thing that I
want you to remember.

63
Jari Arkko 00:25:01.394 --> 00:25:21.464
So, um, we're trying to take an end to end view across devices networks, data
centers and applications. This is not just a big guy Internet, and, you know,
just the transmission of packets in the routers. But but it's the whole thing.
Um, we should remember that the impacts can come from, you.

64
Jari Arkko 00:25:21.644 --> 00:25:42.404
Of course, but also manufacturing and decommissioning what kind of, uh, E,
waste remains um, we should worry about greenhouse gases obviously. Um, but
also other issues might might be quite relevant. Uh, we consider both 1st and
2nd order impacts. So, um, you know, 1st, order impact is.

65
Jari Arkko 00:25:42.765 --> 00:26:03.735
Is that we, uh, for instance, produce some pollution or from running our
systems but then there's the 2nd order that, like, well, we can, for instance,
help the, the society to do something, you know, more efficient manner avoid
flying for instance or some sense. Um, of course, those impacts can be.

66
Jari Arkko 00:26:03.794 --> 00:26:23.684
Both positive and negative, so we have to remember that, uh, improvements are
obviously interesting for this workshop, but not only that it's not just about
fixing stuff it's support also about understanding stuff. So, if we at least
understand that here is an issue, then then that's, you know, where 1 step
further.

67
Jari Arkko 00:26:25.784 --> 00:26:45.944
And it's also important to understand the trade offs and costs of, you know,
whatever we would change, uh, Alvaro and others, um, commented that on on their
position paper. So recommend reading and we can't do growth. So, small business
and other non technical aspects and or or even regulates and governments. So
those.

68
Jari Arkko 00:26:46.154 --> 00:27:07.214
Sort of linked, um, solutions they can come from many, many different, uh,
angles. It can be about implementations protocols and standards, which are, you
know, for many of us. It's the main thing. Uh, we can also be able to clean
energy can be changes in business practices. You know, imagine, for instance,
if you didn't have advertisement based.

69
Jari Arkko 00:27:07.275 --> 00:27:27.075
Models W, would that impact somehow? I, I don't know because I don't think we
have measurements, but it's an interesting question. Um, we could have better
awareness, better measurements and better transparency in the networks that,
hey, this is, this is what's happening because it'll be a request, or we could
simply be building better tools for somebody else to do.

70
Jari Arkko 00:27:28.395 --> 00:27:28.815
For instance.

71
Jari Arkko 00:27:32.144 --> 00:27:44.264
Um, so that's it from my side, if there's, um, any, um, quick comments on this,
we can take that otherwise we can proceed to.

72
Jari Arkko 00:27:49.124 --> 00:27:52.154
And are you online and are you able to share.

73
Jari Arkko 00:27:57.045 --> 00:27:58.455
Yep, thank you.

74
Toerless Eckert 00:28:15.344 --> 00:28:23.444
Okay, small fight with the unmute button and classic. All right. Can you see
the slides? Hello?

75
Jari Arkko 00:28:23.474 --> 00:28:23.954
Yes.

76
Toerless Eckert 00:28:24.044 --> 00:28:24.434
Okay.

77
Jari Arkko 00:28:24.494 --> 00:28:24.764
Yes.

78
Toerless Eckert 00:28:25.574 --> 00:28:33.224
All right, so let's look into the review mirror. So this is a position paper
on, um, what we have done.

79
Toerless Eckert 00:28:33.314 --> 00:28:37.394
Um, around energy, uh, in the.

80
Toerless Eckert 00:28:37.484 --> 00:28:43.064
And, uh, this intro slide shows who's been working on that and, uh, where you
can find and discuss it as well.

81
Toerless Eckert 00:28:45.194 --> 00:29:05.654
So, why did we start this work now? So last year, um, IP, RFC, 70, 91 turned 40
years old and we didn't have a party. Of course, there was covered, so maybe
time for an early mid life crisis. Um, so there's a lot of positive things
obviously we're aware of what our technology has done. Right? And I think that
became particularly.

82
Toerless Eckert 00:29:05.684 --> 00:29:26.804
You're in Corona, when nobody could physically move and, uh, it was really the
Internet and other tcp networks that, uh, enabled, you know, for society to
really continue to, to operate much better than it would have been possible 30
years ago or 1018. I think it was, um.

83
Toerless Eckert 00:29:27.074 --> 00:29:47.864
But, on the other hand, uh, 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. Um, so before looking forward, it seems good to
start analyzing what we've done.

84
Toerless Eckert 00:29:48.464 --> 00:30:08.684
And, uh, the hope of course, is that it enables more contributors for future
work to, um, understand where we are and what we have done, um, find, uh, in
network, uh, gaps. But, of course, as, uh, people with cool, uh, slogans have
told us for hundreds of years. Um, just because we know what the past is
doesn't mean that.

85
Toerless Eckert 00:30:10.065 --> 00:30:30.255
We know exactly what to do next. Um, so, what's the scope and how do we go go
along right? So, um, especially amongst my colleagues, um, some of them, like,
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.

86
Toerless Eckert 00:30:30.284 --> 00:30:51.284
Cost of the network infrastructure versus the client data center, mobile use
cases. So it's it's really overwhelming if you think about the impact of energy
in and society, uh, around ID. Um, so of course, it became a lot easier when
I'm looking at this document by simply saying, well, we'll have to somehow
prove what the.

87
Toerless Eckert 00:30:51.704 --> 00:31:12.404
With it by relating it to, um, the work that the has done, which is finished or
also attempted and, uh, as of yet abandoned drafts. Um, and then basically try
to, um, structure, um, the work around that that makes the document a little
bit, uh, hard to reach, uh, with all the references in it. But obviously, for
people.

88
Toerless Eckert 00:31:12.794 --> 00:31:33.614
That we're targeting, like, the ones, uh, hopefully doing new work with it. I
think that's the best that we can do. And it turns out that most of the impact
that our work has done, um, is really incidental. Right? Because energy often
became a relevant metric of considerations much later than the work itself. Um,
an intentional energy.

89
Toerless Eckert 00:31:33.734 --> 00:31:54.854
Work only recently happened in the last, maybe 20 years right? And, um, 1 of
the other points was, uh, you know, to recognize that we need to expand from
energy to sustainability by using carbon based metrics, uh, whenever the use
case, and workflows of the technology. Uh, really allowed to take that.

90
Toerless Eckert 00:31:54.885 --> 00:32:15.975
To consideration, um, because that pretty much means that the same energy
consumption is less problematic, uh, if it can use renewable energy, then if it
would have to use carbon burning energy. All right, so here is just a single
page overview of what the document, uh, entails its 40 pages. So, um, no, I
hope.

91
Toerless Eckert 00:32:16.034 --> 00:32:37.154
Many of you will have read it or will read it after this, uh, um, these slides.
Um, it starts with really the, um, high level tenants of the Internet and other
tcp networks of the architecture networks. Right? And that is in general energy
saving through scale. Right? And there are a lot of technologies.

92
Toerless Eckert 00:32:37.364 --> 00:32:58.304
Produced scale through convergence, um, through global networking through
federation, you know, the Internet itself is the biggest Federated network.
There is encryption and freedom to innovate. Right? All these things brought
more applications more users, more work flows into the same networks, allow
them to scale.

93
Toerless Eckert 00:32:58.394 --> 00:33:19.304
Cost, and therefore make the solutions even more competitive and attractive,
um, uh, then, you know, their prior, um, alternatives, if they then existed and
I'm going into, um, especially details of applications we are mostly aware of,
like the Tele, collaboration where we've seen, you know, a 40 year history from
email through telephony.

94
Toerless Eckert 00:33:19.574 --> 00:33:40.604
And then, uh, lately, web, which is at the core of all the video conferencing
that, you know, has I'd say a safety plan is communications during corona
times. Right? But, of course, the more, uh, uh, broader topic of digitization,
replacement of work flows without it with those, uh, that use the Internet or
other networks.

95
Toerless Eckert 00:33:40.905 --> 00:34:01.755
Is, uh, also captured there, um, then exactly that expansion from, um, energy
saving as, um, the, the core benefit to, to enabling the use of sustainable
energy um, and computation with that for example. So, uh, I give some examples
of how, um, the Internet and technology.

96
Toerless Eckert 00:34:01.784 --> 00:34:22.904
Uh, benefit there, and then ultimately we get into the area where energy was
exactly 1 of the core targets of improving the protocol landscape. And that is
what we call the low power and glossy.

97
Toerless Eckert 00:34:22.909 --> 00:34:43.845
Train networks, and we have a humongous set of, uh, working groups that have
been working on that and it all comes from devices with very little ability to,
um, have energy whether it's battery based, or self production of the energy.
Um, so there is, um, um, really good, uh, great amount of work, which I think
will also proliferate.

98
Toerless Eckert 00:34:44.594 --> 00:35:05.204
Into faster, uh, and, uh, more energy available networks in the future, just
because they're all more efficient than their prior counterparts. Um, then
finally, um, a lot of technology details with simple technologies that save
energy, sleepy nodes multicast, um, discussions around, uh, the use of
technology.

99
Toerless Eckert 00:35:05.209 --> 00:35:26.234
In the energy networks themselves, the smart grid, the central phaser networks,
which enable really to avoid brown outs on the planet much better than in the
past working groups for measurements that we had. Um, and, uh, metrics and
benchmarking and then the power awareness of these metrics.

100
Toerless Eckert 00:35:26.415 --> 00:35:47.505
Which, you know, hasn't gone very far, uh, today via dependent effort. And, um,
uh, ultimately, of course, what's happening today is through mechanisms. So
that's kind of the, the rundown. Um, I haven't counted. I, I think it's in
excess of 100 references to and drafts. That are in the documents so, um, there
is a human.

101
Toerless Eckert 00:35:47.534 --> 00:36:08.384
This amount of work that we've done at least when you start thinking about how
the energy related impact of work um, specific documents is not only the
networks that were produced with them. All right. So, um, reviewers asked me.
Okay. So, rearview mirror is nice. But, uh, what really is next, and I was
trying to figure out here 3 points, right?

102
Toerless Eckert 00:36:09.675 --> 00:36:29.805
So 1st of all, I think they're a significant, but not easy, optimization and
enhancement at the network layer. Um, I think there are a lot more, you know,
uh, workflows as we've seen with web latest and others that I think, uh, if
we're engaged in the, uh, workflows using the network, um, they, they can
become better. Um, but.

103
Toerless Eckert 00:36:29.809 --> 00:36:50.954
Ultimately, when we look at the inflection points, the the really big changes
in our history of the Internet, I think we have the Internet itself mobility
data centers as maybe the top 3. and I think the next already starting big
inflection point is renewable energy. Because it will break barriers for energy
consumption. Maybe only decades out, but we're.

104
Toerless Eckert 00:36:50.984 --> 00:37:11.954
Really seeing the starting in pockets um, I think it's going to go through all
forms of energy consumption. Um, and many of those will need ity components for
that. Um, and, uh, I've, I've been giving a few examples, like the time and
place shift, but ultimately, when there is a surplus of energy, what are the
next.

105
Toerless Eckert 00:37:12.375 --> 00:37:20.535
You know, it, Internet based, um, uh, use cases workflows that those would
enable. All right and that's it. Thank you.

106
Jari Arkko 00:37:27.735 --> 00:37:34.035
Thanks and, uh, we do have time for questions if any otherwise, um, we can let
eve continue.

107
Jari Arkko 00:37:38.774 --> 00:37:44.174
Any questions in the chat, at least so moving forward you can you share.

108
Eve Schooler 00:37:45.314 --> 00:37:52.454
Sure, Chris, were you coming on video to ask a question? Okay. Okay. Let me
share my screen.

109
Eve Schooler 00:37:59.564 --> 00:38:01.214
Are you seeing that.

110
Jari Arkko 00:38:04.725 --> 00:38:07.875
See your screen yes, thank you.

111
Eve Schooler 00:38:07.875 --> 00:38:12.735
Let's see trying to get this to go away.

112
Eve Schooler 00:38:16.005 --> 00:38:19.425
A lot of so you're probably seeing the pop ups as well.

113
Eve Schooler 00:38:21.404 --> 00:38:21.734
Sorry.

114
Rick Taylor 00:38:23.024 --> 00:38:25.214
No, we're not seeing the pop ups. That's that's good.

115
Eve Schooler 00:38:26.024 --> 00:38:29.414
Okay, let me are you seeing.

116
Rick Taylor 00:38:30.194 --> 00:38:30.674
Full screen.

117
Eve Schooler 00:38:34.214 --> 00:38:37.004
I'm a little confused here, but, um.

118
Eve Schooler 00:38:38.145 --> 00:38:44.475
A lot of stuff popped up on my screen. So, are are you seeing the, the notes?
Are you seeing the.

119
Rick Taylor 00:38:44.775 --> 00:38:47.745
Just the, just the slides just the slides full screen.

120
Eve Schooler 00:38:48.105 --> 00:38:48.375
Oh.

121
Jari Arkko 00:38:48.915 --> 00:38:51.375
2nd flight and nothing else is possible.

122
Eve Schooler 00:38:51.465 --> 00:38:59.265
Okay, so it's kind of reverse to my screen so that's why I'm a little bit
confused. Okay. But anyway, uh, suffice to say I'm a Schuler I'm from Intel and.

123
Eve Schooler 00:38:59.384 --> 00:39:20.414
I have done this work with my colleagues from, uh, Oxford and Ari and Yale and
Intel and, uh, we Co, authored the position paper perspective on carbon aware
networking and, uh, the request for this lightning talk, however, was really to
focus on carbon versus energy.

124
Eve Schooler 00:39:20.564 --> 00:39:41.564
And that's really because, um, if you look at sustainability research within,
uh, community, the information communication, technology, community or sector,
it's traditionally focused on energy efficiency and the goal to consume less
energy.

125
Eve Schooler 00:39:41.624 --> 00:40:02.714
However, there's this complimentary goal, which is really to decarbonise the
energy consumed. So if you're going to consume energy, ensure that it's green,
it's generated by renewables or other clean energy technologies. And then
really, we consider that there's this 3rd pillar for sustainable, which is to
minimum.

126
Eve Schooler 00:40:03.105 --> 00:40:23.865
Um, other environmental impacts, um, and today, and for most of this workshop,
I think we were largely here about energy decarbonization. Um, the effective
the most effective action, of course, is, um, at the intersection of this Venn
diagram and more typically, or I would say, paradoxically, it's.

127
Eve Schooler 00:40:23.895 --> 00:40:44.745
In the case that 1 pillar is achieved, but not necessarily the others, so for
for example, 0, emissions does not necessarily equal energy, efficient case in
point being some of what's going on with renewable powered Bitcoin mining and
cryptocurrency data centers, nor does energy efficiency equals 0. emissions.

128
Eve Schooler 00:40:45.434 --> 00:41:06.164
A high energy efficiency, uh, factory may be located in geography where the
power is generated from fossil fuel and therefore has carbon high, carbon
intensity. Um, and so, uh, it really again, I'm struggling a little bit with my
screens here. But, um, so.

129
Eve Schooler 00:41:06.194 --> 00:41:27.284
So, to really reduce the growing electricity usage of the sector, and in turn
the carbon footprint of data centers, some of the hyper scalers have developed
and deployed what they are calling either carbon, intelligent or carbon aware
computing. And that's really.

130
Eve Schooler 00:41:27.344 --> 00:41:48.404
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 data centers and.

131
Eve Schooler 00:41:48.470 --> 00:42:09.555
General to lower their current footprints and in turn helps stabilize the grid
by maximizing the usage of renewable energy and consuming as tourists referred
to excess energy that would otherwise go unused. And in these geographies like
California and Germany where renewables integration and.

132
Eve Schooler 00:42:09.620 --> 00:42:30.765
Energy generation from neural renewables is outpacing consumption. Um, uh,
that's a really important stabilizing aspect for the grid. So this carbon
awareness allows data centers to load balance the grid or to service virtual
batteries. And our paper refers to some of the references for that.

133
Eve Schooler 00:42:31.064 --> 00:42:42.044
Um, but it more importantly leads us to ask, why not employ carbon awareness
everywhere and throughout the entire ecosystem ecosystem.

134
Eve Schooler 00:42:42.049 --> 00:43:03.194
And not just in hybrid hyperscale data centers, but throughout the edge to
cloud continuum, um, throughout as we begin to roll out next generation
infrastructure, um, particularly in places where it's possible to more
approximately Co locate with renewables, uh, let's not just focus on carbon.

135
Eve Schooler 00:43:03.224 --> 00:43:24.314
Compute, but other facets of systems, like storage and networking and begin to
adopt carbon and test intensity as 1 of these quality of service metrics. And,
um, we of course want to look at not just what can we do with hardware, but all
the way through software components for the up and down.

136
Eve Schooler 00:43:25.215 --> 00:43:45.495
Um, and in in what software that's resident in a particular platform, but also
end to end horizontally and we are going to need a lot of entry points in the
architecture for APIs that allow us to, uh, gain some insights from carbon
awareness. I've included a couple of graphs on the right and, uh, they're blown
up.

137
Eve Schooler 00:43:45.524 --> 00:44:06.644
In the background on my slides, so you can prove 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. Um, the, the orange in that is the
data center growth, but you can see there's.

138
Eve Schooler 00:44:06.650 --> 00:44:26.325
A lot that lies outside of the data centers and, uh, when you study the
details, you can see that networking is definitely on par with, um, data
centers, in terms of its carbon footprint. Um, so there's some self reflection
we should be doing, but the, the bottom graph is actually from, uh, California
is.

139
Eve Schooler 00:44:28.244 --> 00:44:40.544
Uh, 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. Um, and you
can see the variability and growth of that as well.

140
Eve Schooler 00:44:42.435 --> 00:45:03.015
Um, and so it leads us naturally to, uh, what my carbon, where networking look
like from an AP perspective. So, uh, again, because this is a lightning talk,
I'm gonna sort of quickly go through this, uh, for carbon aware routing. We'd
really like our current routing protocols to be able to select.

141
Eve Schooler 00:45:03.344 --> 00:45:24.164
We'd like them to evolve to be able to select more carbon efficient paths,
possibly by considering carbon intensity as an additional metric. But also, by
acknowledging that we have time variant links some, that might be powered by
renewable energy. And that's my predictably becoming and going. Um.

142
Eve Schooler 00:45:24.284 --> 00:45:45.344
Inspired by some of the delay tolerant network, uh, activity in there's and the
intermittent of green energy we'd also like to construct some carbon aware
transports to schedule the time and space shifting of data transmissions. Not
just orchestrated workloads. Um, and to minimize the carbon cost.

143
Eve Schooler 00:45:45.350 --> 00:45:56.625
For use cases that are time elastic naturally such as bulk data transfers for
backups for software updates for telemetry data collection, and so forth.

144
Eve Schooler 00:45:58.039 --> 00:46:18.704
Inspired, I think by some of the work going on in the detonate or deterministic
networking community carbon where traffic engineering could be called upon to
guarantee that flows, stay within carbon consumption budgets, possibly
employing hub. I hop techniques.

145
Eve Schooler 00:46:19.214 --> 00:46:34.304
And, uh, possibly also reserving clean energy resources along the way. And, of
course, foundational to everything is that carbon telemetry, uh, needs to exist
for all this work. And it requires network element observed ability.

146
Eve Schooler 00:46:34.904 --> 00:46:38.894
But also an awareness and reduction of its own impact.

147
Eve Schooler 00:46:42.495 --> 00:47:01.485
And, um, again, due to time here, W, will queue up, you know, what the
challenges are here for discussion in the session for. We're gonna have another
lightning talk to talk about the next steps for carbon aware, um, networking.
Um, but suffice to say here that.

148
Eve Schooler 00:47:01.724 --> 00:47:22.244
Some of the challenges are really to make real time, um, tracking and reporting
for electricity consumption. Something that's, uh, accessible additionally
tracking and reporting for finer grain, electricity, carbon intensity and those
things together mapped to.

149
Eve Schooler 00:47:23.084 --> 00:47:43.394
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 a, and and, um,
obviously.

150
Eve Schooler 00:47:43.994 --> 00:47:52.484
Once you have all that information, what is it that we're going to actually do
with it? How are we going to react in near time near real time to carbon
related information?

151
Eve Schooler 00:47:55.514 --> 00:48:15.884
Uh, uh, this is just to point you had a couple other readings if you want to
look at for the details but then there's some terrific, um, papers throughout
the workshop. Um, having to do with, uh, carbon in general, uh, whether that's,
uh, paper on pricing, um, of, uh, what's going on.

152
Eve Schooler 00:48:16.305 --> 00:48:37.365
Behind the scenes, the sort of business elements there's, um, additional,
there's another paper on carbon Ware networking, um, the Chris Adams uh, and
his Co authors wrote as well as, uh, interesting papers on metrics. That
clearly pertain to this as well as, you know, power proportionality. And how
that affects, um, energy.

153
Eve Schooler 00:48:37.395 --> 00:48:51.405
And in turn the carbon impact, as things are both static and variable. So there
you have it. That was my I talk, I don't know how to stop sharing.

154
Eve Schooler 00:48:54.464 --> 00:48:55.454
Let's see. Oh, I think.

155
Jari Arkko 00:48:55.454 --> 00:48:58.514
Maybe the top of your screen and then stop. Yeah.

156
Eve Schooler 00:48:58.514 --> 00:48:59.234
Here you go.

157
Jari Arkko 00:49:00.044 --> 00:49:14.714
Yeah, good. Um, thank you. That was very interesting. Uh, I think I will at
least have some questions. There's also some other questions on the chat. Um,
but maybe we'll go to best stuff. 1st, are you online and can you share you
also have a.

158
Jari Arkko 00:49:15.464 --> 00:49:18.224
And very interesting angle. There we go. Thank you.

159
Vesna Manojlovic 00:49:20.534 --> 00:49:24.164
Hi, everyone, I'm, I.

160
Jari Arkko 00:49:24.224 --> 00:49:31.694
You are your voice level is pretty low if you can adjust otherwise we'll tune
up uh, our Williams.

161
Vesna Manojlovic 00:49:32.054 --> 00:49:33.584
I'll try speaking louder.

162
Jari Arkko 00:49:34.244 --> 00:49:35.234
That's much better. Thank you.

163
Vesna Manojlovic 00:49:37.005 --> 00:49:57.705
I work as a community builder for however, today I'm speaking as an activist,
and as an intersectional feminist, and as a parent of activist. So I submitted
the paper that have been seen very, very political. So I've.

164
Vesna Manojlovic 00:49:57.764 --> 00:50:18.884
Us to talk more about the, uh, societal aspect of where does the environmental
sustainability fit um, not only from the technical perspective. And, uh, so I
wish to propose that we don't limit ourselves to only talking about.

165
Vesna Manojlovic 00:50:18.914 --> 00:50:40.034
Environmental sustainability, but work towards climate justice. So I'd like to
on this 1 slide propose like 3 problems and 3 possible solutions in a broadest
possible sense. 1 of the structural problems of our.

166
Vesna Manojlovic 00:50:40.039 --> 00:51:01.184
Our society and politics and economics, rather than just technical is the
inequality of the environmental impact with something that we use to call
digital divide. Um, referring to the lack of access to the digital technologies
in the circle, global South. But now.

167
Vesna Manojlovic 00:51:01.484 --> 00:51:22.334
We can actually also call a digital colonialism in which we are, um, kind of
repeating the history of the actual economic colonialism in the digital sphere
where we copied the exploitations economic exploitation and nature exploitation.

168
Vesna Manojlovic 00:51:22.814 --> 00:51:43.364
Uh, that are predominantly, uh, kind of externalized 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.

169
Vesna Manojlovic 00:51:43.489 --> 00:52:01.964
North, and another part of that is the policies of extra activism, so just
taking as much energy resources, raw materials, water in order to produce the
digital devices. And then later.

170
Vesna Manojlovic 00:52:03.494 --> 00:52:24.374
Kind of throwing them away and, uh, again, predominantly in the global South
order in an in unequal way, the way how to deal with that would be to consider
ourselves as 1 humanity and, uh, to show.

171
Vesna Manojlovic 00:52:24.379 --> 00:52:45.524
Solidarity with the parts of the world, and the communities that have been
already impacted this, proportionally with this development and to at least
minimize the harms. And we, the goal of stopping the harms to those communities
and those natural environments.

172
Vesna Manojlovic 00:52:45.555 --> 00:53:06.525
2nd problem is our belief in the endless growth on 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.

173
Vesna Manojlovic 00:53:06.704 --> 00:53:27.824
Actually 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.

174
Vesna Manojlovic 00:53:27.854 --> 00:53:48.884
And aiming for convenience, and so to combat that, we need to change the way of
thinking and to implement the thinking of sufficiency or even modesty and
frugality in our technical design, considering limited constructivism,
decreasing the.

175
Vesna Manojlovic 00:53:48.979 --> 00:54:09.974
Growth and focusing on the common resources and common goods, or in a more like
down to earth deck on the free and open source software and open hardware and
open standards that the idea is famous for and finally addiction to fossil
fuels.

176
Vesna Manojlovic 00:54:10.129 --> 00:54:30.974
And all the other over consumed materials, like water, land and minerals. We
have to just stop doing that Sir or 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.

177
Vesna Manojlovic 00:54:31.305 --> 00:54:52.365
For the more, um, uh, life affirming needs of the people, rather than for the
data centers and so in the tech field, we should be focusing on repeatability,
circular economy and durable tech. My next slide is adjusted.

178
Vesna Manojlovic 00:54:53.354 --> 00:55:13.574
Translating these suggestions for the, uh, follow up sessions on this
workshops. So, in the 2nd 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.

179
Vesna Manojlovic 00:55:13.604 --> 00:55:34.694
More on the actual activities, 2nd, when we consider the improvement, I'm
suggesting that we consider something opposite from what we were doing until
now. And that is not growing but decreasing because the climate goals are also
saying that we have to decrease year on year.

180
Vesna Manojlovic 00:55:34.729 --> 00:55:55.874
According to the certain percentage, so we could put that in all of our design
plans and be very ambitious and say we want to decrease everything by, let's
say, 10% per year. And the other advice would be to consider that the future is
going to contain.

181
Vesna Manojlovic 00:55:56.234 --> 00:56:16.934
Uh, 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.

182
Vesna Manojlovic 00:56:17.030 --> 00:56:38.055
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 and, uh,
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.

183
Vesna Manojlovic 00:56:38.444 --> 00:56:53.774
Uh, there is already an RFC about that, uh, how the Internet is for the end
users, uh, the research groups, like Gaia and activist, and also the, uh,
existing climate justice organizations.

184
Vesna Manojlovic 00:56:56.834 --> 00:57:03.824
And, uh, there won't be any Internet on the burning planet. So that's why I'm
speaking here as an activist. Thank you.

185
Jari Arkko 00:57:06.675 --> 00:57:25.815
Smart that was also super interesting and a worthwhile angle to to discuss. Um,
so now we have about 15 minutes for discussion. If you want on the queue, um,
just, uh, put plus queue in the in the chat and actually.

186
Jari Arkko 00:57:26.174 --> 00:57:47.264
That, uh, as an example, and also, because I wanted to ask a question. So I'll
shoot my question. 1st, um, Eve inspired by your presentation, which was very
interesting. Um, so you talked about this carbon awareness and is that what
the, the approach is that would that be.

187
Jari Arkko 00:57:47.354 --> 00:58:02.024
Mostly, in fact, domain or inter domain, because I presume that the challenge
is between those 2 would be very different. Like, you can, he needs a domain.
You can trust mostly what the devices say but inter domain that might be much
more difficult.

188
Eve Schooler 00:58:02.354 --> 00:58:02.744
Right.

189
Jari Arkko 00:58:02.984 --> 00:58:04.454
Where is the focus mostly?

190
Eve Schooler 00:58:04.754 --> 00:58:07.904
Right I I don't know where where it will be, but, I mean, you.

191
Eve Schooler 00:58:09.495 --> 00:58:29.565
Targeted, uh, something that we've been thinking about, which is, um, within an
administrative domain, 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.

192
Eve Schooler 00:58:29.594 --> 00:58:50.354
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 not only
that, but there's so many elements that are hidden from you in our network
paths that also consume energy that have these secondary effects.

193
Eve Schooler 00:58:51.135 --> 00:59:11.145
So, is it sufficient to just measure the consumption of energy along a route?
There's like this whole supporting cast that we don't quite know how to, um,
comprehend just yet. And so we are going to have to partner with other, is that
we transit um, and, uh, so.

194
Eve Schooler 00:59:11.894 --> 00:59:33.014
We don't have a solution is yet I mean, I think that's where the research is,
what should be exposed should be, where should it be exposed should be exposed
at the boundaries. So that for hop technology, we can kind of accumulate ranges
of expectation around, um, energy usage and carbon intensities and things like
that. So, it's a big unknown. And so we.

195
Eve Schooler 00:59:33.104 --> 00:59:40.994
People's thoughts either off the cuff, or if they have experience in things
that are similar to this for example.

196
Jari Arkko 00:59:42.764 --> 00:59:48.284
Thank you looks like the researches will have some work work to do, uh, on the
queue. That's Lars.

197
Lars Eggert 00:59:50.624 --> 00:59:53.834
Hello. Um, so this is really interesting.

198
Lars Eggert 00:59:54.170 --> 01:00:15.165
Quite diverse, so I actually sort of felt that that's not sort of summarized a
bunch of points that I've sort of been bouncing around in my head. Right?
Because, you know, obviously we can try and build an Internet it uses like,
half the power right? Do we rip all the equipment out and we put new equipment
and then, um, you know, then we, we're great. We're only using half the power.
But but in terms of.

199
Lars Eggert 01:00:15.734 --> 01:00:36.434
And balance we've probably done pretty poorly because we, you know, just
doubled the embodied, uh, carbon of the Internet. And so I think we gotta get
to the point that, that that's not what's lining out. And she did it in a very
active its way. But I really feel sympathetic to it because we need to sort of
not look at the Internet and its capacity is unlimited. Right? We've always
sort of been banking on the Internet.

200
Lars Eggert 01:00:36.470 --> 01:00:57.585
Getting better over time and and it does that by using more and more resources
and I think we got to 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
where we don't generate the data. The data gets generated by various services.

201
Lars Eggert 01:00:57.644 --> 01:01:11.924
Platforms and what have you, um, but I think this sort of notion that, you
know, um, 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.

202
Jari Arkko 01:01:14.894 --> 01:01:19.094
Uh, you go ahead and after you, uh, it will be and.

203
Jari Arkko 01:01:26.564 --> 01:01:27.524
We're not hearing you.

204
Jukka Manner 01:01:29.145 --> 01:01:49.755
Sorry, user error uh, 1 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, uh.

205
Jukka Manner 01:01:50.534 --> 01:02:10.034
Kind of claiming that if we are using renewable energy, then all is fine. But,
for example, I was talking to 1 major cloud service provider a few a few months
ago. And it was interesting that, you know, they plan on having everything
based on renewal energy by 2030.

206
Jukka Manner 01:02:10.964 --> 01:02:15.224
Then I asked that what is their absolute energy consumption going to look like.

207
Jukka Manner 01:02:16.604 --> 01:02:37.664
The answer was that it grows by 25% per year we're talking about give or take
500% of more renewable energy needed to be built for this 1, single player. So,
renewable doesn't mean carbon free and that's the point that, you know, there
in my opinion.

208
Jukka Manner 01:02:37.729 --> 01:02:58.874
To answer that 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
best numbers when I was saying, and there's many many topics. I agree in the
big picture related to this whole.

209
Jukka Manner 01:02:58.879 --> 01:03:05.564
T, and consumption and so forth, but maybe I'll stop here. Otherwise I'm going
to give a lecture and that's not the point today.

210
Jari Arkko 01:03:08.474 --> 01:03:10.724
Right Thank you. And then so.

211
Suresh Krishnan 01:03:11.264 --> 01:03:28.484
Thank you thanks. You like the so, that was really fascinating presentation.
So, 1 question I had is, like, you're talking about how buy hop stuff um, do
you know of anything that, like, you know, how these, like, hops are a portion,
their share of, like, carbon emissions like, is there something um, we can
actually do that.

212
Suresh Krishnan 01:03:28.514 --> 01:03:38.624
Right because it seems like a very difficult problem. Like, we never managed to
do this, like, in a multi domain basis, or multi administered the domain basis.
So, is there any thoughts like how you would go about it?

213
Eve Schooler 01:03:41.354 --> 01:04:00.884
I, I was drawing analogies to, um, sort of what happens in the measurement
community as you transit paths, and you collect information measurement
information. Um, and so, you know, I don't know whether that those might be the
same kinds of technologies that we, we would use your protocols that do that,
but.

214
Eve Schooler 01:04:00.920 --> 01:04:22.005
Right. We haven't traditionally looked beyond specific kinds of attributes
about our network performance, like packet loss and latency and the variance
and the latency of the residents jitter and it begs the question. Can we even
rely on external.

215
Eve Schooler 01:04:22.069 --> 01:04:43.064
Measurement from other places to tell us about the energy consumed or the
carbon intensity, which are things that might fall outside, uh, the ownership
of those who are operating the networks and so we need to have some kind of
trust relationship there. I don't I can't say that.

216
Eve Schooler 01:04:43.725 --> 01:05:03.885
Um, you know, which technologies have really scaled in order to do hop 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 is 1 of these protocols that
they did some of that there there are other things out there. I was thinking
more in the small scale.

217
Eve Schooler 01:05:04.664 --> 01:05:25.514
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 that might be places
to glean for insights there, whether it's worked or not at scale. Um, I need to
probably.

218
Eve Schooler 01:05:25.544 --> 01:05:29.234
To those more on the measurement side.

219
Suresh Krishnan 01:05:29.264 --> 01:05:30.134
Thank you thank you.

220
Jari Arkko 01:05:30.884 --> 01:05:31.904
Thanks vanilla.

221
Pernilla Bergmark Ericsson 01:05:35.984 --> 01:05:53.714
I oh, sorry I couldn't join you from start. I'm penny. Lima colleague, Jerry,
and I, uh, work as a principal researcher for sustainability impact and I'm
also, uh, Co, chairing the standardization work in ICU on the development of,
uh.

222
Pernilla Bergmark Ericsson 01:05:53.745 --> 01:06:14.865
Standards, uh, so I just wanted to 1st, I've I'm very glad to be here. It seems
like a very interesting event. Uh, but I also wanted to to both comment on what
you said, and also to refer back to the work that we have done in, because that
could be be of interest here.

223
Pernilla Bergmark Ericsson 01:06:14.894 --> 01:06:35.864
So, I agree with W. W. so in in the frame of I to you, we have developed a
trajectory, which is a normative, uh, and feasible trajectory for the sector.
And when we talk about the sector, we talk about, uh, data center networks and
use the device. Is taken from a life cycle perspective.

224
Pernilla Bergmark Ericsson 01:06:36.254 --> 01:06:56.534
And it says that the sector should, uh, reduce its emission by, uh, the
footprint emissions by 45% from 2020 to 2013. this has also been adopted by the
science based target initiative. It was developed jointly by science, based
toast initiative. Iq DSM may Jesse.

225
Pernilla Bergmark Ericsson 01:06:57.285 --> 01:07:18.225
We had also participation from the, um, so, uh, what what we have seen in this,
uh, trajectory, is that about 80% of the overall lifecycle footprint, uh,
relates to the use of electricity and so, of course, eh, switching to renewable.

226
Pernilla Bergmark Ericsson 01:07:18.764 --> 01:07:38.174
Um, per month, it's very important, but it's not sufficient. Um, 1st, it's not
the full footprint. It's not the full, uh, perspective of environmental impact
and it's also, uh, as you mentioned, uh, it's.

227
Pernilla Bergmark Ericsson 01:07:39.555 --> 01:07:59.805
We still need to produce the, even if we come from from renewable sources and
this will be, of course, uh, um, a constraint in the availability. So there are
many reasons to to, to think broader. Uh, but but this, this, uh, trajectory we
have developed to gather with.

228
Pernilla Bergmark Ericsson 01:08:01.034 --> 01:08:06.464
Standard, uh, which is aligned with, uh, such as the, uh, um.

229
Pernilla Bergmark Ericsson 01:08:07.995 --> 01:08:28.785
The race to and the science based targeting initiative, that standard. Uh, but
from the assessment perspective, we have in particular developed an standard
for, for products networks and services. There is a, a standard for calculating
the, uh, footprint of the entire sector. And what shouldn't be.

230
Pernilla Bergmark Ericsson 01:08:28.844 --> 01:08:49.963
Taken into account and so on and this is important because we see that many
startups are quite simplistic. Let's say, and they make assumptions on models,
which are just scale with data, though. We know that the majority of that to
keep this.

231
Pernilla Bergmark Ericsson 01:08:49.995 --> 01:09:11.115
Systems up and running and so on. Um, and then we have also exactly today
actually, uh, ex agreed, uh, or approved even a standard, which is dealing with
the indirect, uh, how to to quantify the indirect effects of. So, just for
those were some, something.

232
Pernilla Bergmark Ericsson 01:09:11.120 --> 01:09:14.145
That could be useful to know about. Thank you.

233
Jari Arkko 01:09:18.404 --> 01:09:36.464
Thanks and, um, at least on my scan, I don't see other people we're about to
run out of time. Um, so if people have quick comments, we can take them now
otherwise we could finish and then return to the topics on.

234
Jari Arkko 01:09:36.524 --> 01:09:39.704
Thursday, I did want to say that, um.

235
Jari Arkko 01:09:41.204 --> 01:10:01.874
Uh, like this, you know, this discussion that we just had about, uh, renewable
energy versus, uh, reducing the energy consumption and so, and I think that's,
uh, yeah, probably a trap that we should not fall into too much. It's not a
question of, you know, who solution can solve this whole problem. If we had a
solution like that, you know, it'd be.

236
Jari Arkko 01:10:02.264 --> 01:10:22.304
What's better? We don't actually, we actually have to use all of these tools
that we have have been discussing to lesser or 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. But, um, but it's not just 1 tool. It's, uh.

237
Jari Arkko 01:10:23.055 --> 01:10:36.825
Yeah, clean energy and reductions and better implementations, better standards,
taking into account the societal issues and justice and everything else. So
that that's my conclusion about.

238
Jari Arkko 01:10:38.714 --> 01:10:42.194
The last words, or or or comments.

239
Lars Eggert 01:10:43.304 --> 01:10:54.734
I just wanted to point out that people don't need to wait until Thursday.
Right? There is a mailing list, and we can continue to discussion with in the
chat now on on that list. So, you don't get bored over the next few days when
Nothing's happening in terms of.

240
Jari Arkko 01:10:56.264 --> 01:10:59.564
Yeah, it did and, uh, part of the reason will be.

241
Jari Arkko 01:10:59.714 --> 01:11:11.744
Have separate sessions on separate cases that we can have time between systems
to think about stuff and, you know, reflect and maybe send email and also
prepare for the next next discussions.

242
Jari Arkko 01:11:16.004 --> 01:11:34.124
Okay, then, uh, thank you, it's been a very interesting session with many
different angles and and could discuss and so this clearly work to be done in
terms of, uh, having right. Measurement data. Um, we had some discussion of
that on the chat as well and and, uh, and then the actual improvements. So.

243
Jari Arkko 01:11:35.024 --> 01:11:36.974
Let's get to that Thank you.

244
Pernilla Bergmark Ericsson 01:11:43.034 --> 01:11:44.714
Thank you so much by now.

245
Nina Lövehagen 01:11:45.044 --> 01:11:45.554
Thank you.

246
Ali Rezaki 01:11:46.784 --> 01:11:48.254
It's very much bye. Bye.

247
Eve Schooler 01:11:49.334 --> 01:11:49.664
I think.

248
Vesna Manojlovic 01:11:51.404 --> 01:11:51.764
Bye.

249
Per Andersson 01:11:55.754 --> 01:11:56.444
Thank you. Bye. Bye.