WEBVTT

00:00.000 --> 00:10.560
Okay, it's now in the morning, again, we can start, can you please take a seat?

00:10.560 --> 00:15.560
Welcome everybody, first of all, I would like to thank you for being here so early.

00:15.560 --> 00:19.560
Much appreciated, it's Sunday and in the morning everybody's probably tired and hungover

00:19.560 --> 00:22.000
but you made it.

00:22.000 --> 00:25.840
Welcome to our nevroom, this is the Confidational Computing Devroom, as you probably

00:25.920 --> 00:30.480
know already and before we start, I would like to give you a quick introduction of

00:30.480 --> 00:35.520
the schedule and Confidational Computing in general and our people of the organizer itself.

00:35.520 --> 00:41.920
So I am Ilaria, as you probably figured out and we also have two other Devroom manager

00:41.920 --> 00:47.600
here, one is Paul and then there is your who happens to be probably on the bus right now,

00:47.600 --> 00:52.960
so he will make it here very, very soon, hopefully and then there are three more people

00:53.040 --> 00:58.400
who are unfortunately not here because they could make it for personal reasons, but regardless,

00:58.400 --> 01:03.440
I would love to give them a very big shoutout in case they watch the recording and so there

01:03.440 --> 01:11.840
would be free since Stefan and Fabiano and moving on, just a very quick introduction of Confidational

01:11.840 --> 01:16.800
Computing because we want to keep this beginner friendly as well, so there are many definitions

01:16.800 --> 01:21.840
of Confidational Computing but here I am going to use the one from the Confidational Computing

01:22.800 --> 01:28.320
which is this organization under the Linux Foundation that brings together vendors,

01:28.320 --> 01:33.120
the cloud providers and software developers to accelerate and standardize the user computer

01:33.120 --> 01:41.280
computing and the third definition is actually the protection of data in use by performing

01:41.280 --> 01:48.000
computation in a hardware base that has the trusted execution environment. So key phrase here is data

01:48.000 --> 01:52.080
in use because we know already how to protect data and rest, we can use encryption or data over

01:52.080 --> 02:00.880
the network with good break, all good, all good, all good. So I was saying if we want to protect

02:00.880 --> 02:05.760
data and rest, the data over the network we can use encryption or TLS, but we also want to protect

02:05.760 --> 02:11.280
data in use because if we are processing the data it is generally encrypted which means that

02:11.280 --> 02:18.960
an admin and server can actually see the data. Confidational Computing is targeting exactly

02:18.960 --> 02:25.120
this gap so it protects data while it is being processed by isolating part of the memory

02:25.120 --> 02:30.480
in a trusted execution environment. This means that memory inside the trusted execution environment

02:30.480 --> 02:37.120
is encrypted and even the US and the hypervisor are not able to access it. Of course this requires

02:37.120 --> 02:42.400
a shift in trust so we don't have to trust the cloud provider but we have to trust the hardware

02:42.400 --> 02:49.680
now. And of course we can also increase this trust by attestation. Now on the left there are

02:49.680 --> 02:54.880
some common properties and on the right here we have context dependent properties. So the common

02:54.880 --> 02:59.840
properties across all the trusted execution environments are confidentiality, integrity, data integrity

02:59.840 --> 03:05.040
in code integrity and then we have some other properties that are not provided by all platforms.

03:05.040 --> 03:11.280
For example, code confidentiality, alternative launches, programmability, attestability and recoverability.

03:12.080 --> 03:16.960
And now to give you a quick overview of the timeline of the technical evolution of Confidational

03:16.960 --> 03:23.120
Computing. It started around 2015 with workload isolations within TLS GX which is one of the

03:23.120 --> 03:29.520
first technologies that was invented and then it shifted towards virtual machine isolation with

03:29.520 --> 03:38.400
SIV, TDX, CCI and so on. And the current phase is actually about accelerators so GPUs,

03:39.280 --> 03:44.880
full system rock solutions and so on and the other a different perspective in the same timeline.

03:44.880 --> 03:50.240
Of course Confidational Computing started as an academic project and then slowly became a

03:50.240 --> 03:56.240
false project. And now of course it's also a matter of adoption. So large scale read world

03:56.240 --> 04:03.440
deployments. Of course this come with new set of technical challenges. But so feasibility is

04:03.440 --> 04:08.160
done already. So now it's also a matter of improving reliability of Confidational Computing

04:08.160 --> 04:14.640
and putting it in infrastructure. Now a quick overview of how the Devram is going to look

04:14.640 --> 04:19.280
like today. So we have a lot of talks. So thank you everybody for submitting and thank you so

04:19.280 --> 04:25.600
much for all our speakers. And our talks are going to encompass both firmware, hypervisor,

04:25.600 --> 04:31.920
technologies, hardware platform and last but not least at the station. And if you want to take

04:31.920 --> 04:37.040
a deeper look on our schedule this is how it's going to look like. I just want to I want to

04:37.040 --> 04:41.760
to notice that there will be a five minutes between breaks so you can come and go and the speaker

04:41.760 --> 04:46.160
can set up. I don't think we will do introductions of the speaker. So if you want to speak just

04:46.160 --> 04:55.360
come with your laptop and put set it up. And yeah we will finish at 1245. Our schedule is kind of tight.

04:55.360 --> 05:00.880
So I ask everybody in this room to keep the if you're speaking please keep it in time and if you

05:00.880 --> 05:07.040
have questions and you don't have time for your questions please go to the hallway and move to the

05:07.040 --> 05:13.600
hallway track. And at the inside of course enjoy the Devram. Thank you again for coming and I will

05:13.600 --> 05:19.600
now give five minutes break and give time to the next speaker to come and present. Thank you.

