00:05
Great to see you. I'm hoping this is the last disc you ever see this year. I want to thank you all for coming out. It's a lot of your time, uh which is very valuable. You have a lot of things that are on your plate to do. Uh We have over 1000 customers, over 500
00:27
partners here. We have over 100 press and analysts. It's really, it's really wonderful to see you all and thank you so much for coming out. So, what I'd like to do is um is, is in addition to thanking you for that also, thank you for again awarding us the world's highest uh net promoter score in business. Uh It's really, it's really because of your responses to us working with us closely that
00:54
we're able to provide the kinds of, of uh products that we provide, the kind of service that we provide. I want to thank you. Uh Thank you for this and uh you know, this is the most important number in our company. Uh And it's something that we plan not only on keeping but actually increasing as we go forward. So thanks again, very much.
01:17
So we're talking about everything that you have to do. And in that, in addition to everything you've always had to do, there are all these new challenges in front of you. A I uh to say, uh you know, not to say the least, uh cybersecurity, which, you know, of course, is, uh if anything just keeps getting to be a bigger and bigger
01:33
burden uh every year and then of course, there is application modernization. Now, I don't know, I've been in the business for, I hate to say it uh four decades and it seems like we're always doing application modernization, right? It's just 11 decade after another with uh with new things to modernize. And then finally, you know, now I think is the time where we're gonna make hybrid cloud real
01:56
that is really understand what we need to have on prem what we need to have in the cloud uh and how to create an environment where it's easy to have uh and choose where you put your application and allow it to move between the two environments. Um Now, you know, if we're going to just address all of these different challenges, we can't really do it in this with the same type of thinking that we've always had,
02:23
right? New challenges require new ways of thinking, not just doing the same thing we're doing for a new environment, but rethinking how we do things I've often thought and it's this is ab absolutely true over decades, that it's often my assumptions that get me in trouble. I don't know if you feel the same. It's what I assume to be true.
02:44
That turns out not to be not to be true. Right. So if we look at what are our assumptions, you know, in the data storage market, let's let's really challenging. So, one of those assumptions is that data storage is managed by it and provisioned by it. Part of our assumption about storage is that the storage is,
03:05
is tied specifically to the application stack, it's siloed. Um And the data itself is, is dedicated to that stack. And if we need it somewhere else, we have to copy it. That this, that upgrades like everything else in it is disruptive, you gotta turn it off, applications come down, you put in the new system,
03:25
bring the applications back up again and that storage inherently is inefficient, meaning that you have to buy the amount of storage you need for the next five years and you start using it, you know, over time, but you never really fill it up. Um And it's complex to manage. So these are our assumptions. So let's challenge them.
03:48
What if we could have unified networked storage. In other words, the storage didn't act as individual arrays but acted more like the cloud where you had this unlimited amount of storage available to you. And you only paid for what you used that it was easy to operate as a fleet as a, as a cloud that was auto and automated and that developers had self service so that it wasn't, I didn't have to get involved in every uh
04:17
configuration, every provisioning. Uh but where developers could uh through code, get the type of storage that they needed. What if there was no more scheduled downtime? What if you could have your applications up all the time? Just like what you expect in a SAS environment, right?
04:38
In as SAS environment, you expect the application to be up all the time and of course, in the background, you expect it to be upgraded, updated and constantly improving, but without disrupting your workflow, what if it had the simplicity of SASS at the end of the day? So that's how we're looking at it at pure. Our goal is to build a platform that's consistent, that's as a service and that meets
05:04
all of your storage needs. Not just one particular slice, not just one use case or not just a series of use cases with different platforms, one platform to really satisfy all of the storage needs and make it look as a service and by the way, do it both on the private as well as on public clouds. All right. So this is what we're doing uh with the pure
05:29
storage platform and it starts with Evergreen. Now you might ask why does it start with Evergreen? Well, if you do not build a platform that is inherently non disruptive with its upgrades that cannot be constantly improved or changed without disrupting an application environment. Well, then you, you don't get the fun the foundation of a SAS environment,
05:51
right? So you need the ability to constantly improve the technology without disruption to your developers, to your applications, to your business. It has to always be getting better. So that's the foundation on top of that. Then we've built purity, purity is the one operating system behind all of our products,
06:11
right? Same API S, same management capability. Uh And it's a simple unified infrastructure, it supports all formats. So block file and object, not different operating environments for for different types of storage uh uh structures, it's uh supports both scale up and scale out and it has extraordinarily high density and efficiency.
06:36
So purity is also a foundation because when you operate on one single uh uh operating system, you're able to make improvements across your entire product line. You don't have, you're not fragmenting. In our case, you know, our development team across multiple different um software uh environments, right? And we make it easier for the customers because
07:00
you only have to learn one environment, not multiple. Now, this is something that we've done that's very, very special and new and that is that we've created a control plan. Now, it may be hard to believe, but data storage is the one area in tech that never created a control plane.
07:19
You had a data plane, you had a management plane but no control plane that, that pervades the entire data storage environment. And this is what we've put together. And it's because we had a single operating system that and because we operated on flash, which has high performance that we could do this and that allows you to be able to manage the entire fleet as a fleet rather than as
07:42
individual arrays. It enables you to create for your developers, for your organization, something that looks like a cloud of storage rather than individual arrays satisfying individual use cases. And we've allowed for automated provisioning by policy as well as automated services again by
08:04
policy. And we're going to go into this in greater detail. And then finally, um you know, we do this in uh with the once you put all that together, that is a control plane, the ability to operate as a, as a fleet or as a cloud of storage, single operating environment.
08:21
Well, then you can provide it as a service, right? Complete SAS model. Now, our ser uh our um as a service model is just like in fact, is a SAS model, what does a SAS model mean? But when you buy SASS, you're not being sold or you're not, you don't have specific equipment with serial numbers that you're operating,
08:47
what you get our service level agreements, right? You get a service level agreement that gives you guarantees for how the service will operate, which is exactly what we do with uh evergreen one. It's, it operates entirely like the cloud, it's all through a web interface or um uh or uh um uh through software um interface and we provide consistent
09:14
and constant improvements non disruptively to your environment. So, um you know, the API model, the um the web model is the way that we all want to go in the future. It's what you want to provide to your developers, you don't want trouble tickets and it uh calls what you want is to give them API S to be able to access the services that they need.
09:35
So what we are promising with this model is that we can support all of your workloads, all of the different use cases from terabytes to exabytes from A I to low cost uh backup storage, all on the same environment, uh fleet wide cloud like management and everything in between. So, you know, this is the promise that we're making is one if you will environment to really
10:03
support all of your data storage environments. And this is the day, this is the pure data storage platform, one platform for all of your needs, con consistent uh sim simple evergreen effortless management at scale providing you and as a service experience. So let's take a quick look as to how it performs, how it works because now as opposed
10:29
to array it looks like a cloud of storage. And what does that mean? Well, let's take a standard database app that is it managed because often database uh databases are managed by it rather than the, the the developers. And just like before, it'd be connected directly to its uh to its volumes and those
10:52
volumes might be on an array sitting right next to it. But let's say you need some additional performance that's not available on that array. This system would automatically load balance to other arrays that have greater performance transparent to the application application keeps running. But additional performance or additional capacity is able to be provided um
11:16
automatically by the system based on the policies that you set. Or let's say you have a custom app developed by or in development by your uh developers and they want access to storage. Now today they might go to the cloud. Why would they go to the cloud because they could get it within minutes uh simply through going through a Gooey or an API, that's not something they've been able to do inside the
11:42
enterprise. But with fusion, we offer, you could build custom API S custom storage classes that allow developers to get access to storage capabilities that you have defined based on your corporate uh policies. And by a simple API call, their data storage would be provisioned and and connected to their application automatically.
12:09
I mean this is developer self service, you know, in a private cloud environment, very, very powerful. And that, that storage might be on prem might be Nicolo might be in the cloud all, all determined by the policies that you set depending on the type of application uh environment you want for that. Let's say you have an A I um application and it needs access to data.
12:33
Now today, typically, what you would do is you would copy data um into a data lake, a data warehouse or into a dedicated um uh storage environment for A I. Because the storage that you have in place today was designed specifically for its use case sized both from a performance as well as a capacity standpoint specifically for that uh use case without any extra performance to be able to service data
13:04
for your A I application. But with pure, with the powerful uh with the power of pure flash, we can provide that extra bit of performance and with fusion, it can be networked in such a way that that A I application can get access to the data where it sits rather than having to replicate data all the time into specialized environment
13:25
specifically for the purpose uh of uh of doing analytics. So a very powerful capability that's allowed by the combination of networking and performance uh that we're providing with uh both fusion and in this platform overall. So a very powerful system environment. Now let's look at data at rest at data, at rest there are still things you generally want to do with it,
13:51
you want to tear it, you want to replicate it. Uh There may be, you may want to geo fence it. This type of data cannot leave a certain geographic boundary or a political boundary that is within your own organization. Uh Some data might be able to uh should be viewed by some groups and not by other groups. You can set those policies within fusion so that data stays within,
14:17
it can stay geo fenced and data can be replicated or tiered automatically based on the policies you set. So it's a very powerful offering. All right. Now, let's take a look at why we believe we can do this. We've been providing you um and you've worked with us to create uh products and capabilities that have proven their leadership over 10
14:41
years. We've been 10 years at the upper right of the Gartner magic quadrant. We've delivered products that have lower labor uh 10 X more reliable than everybody else. Uh 2 to 5 times more power, space efficient. So it's, we've proven our leadership over the last 10 years,
14:58
but we are even more ambitious as we go forward as we go forward. We want to lead the way to the next decade of um of, of leadership and storage. In particular, we want automated storage management across your entire fleet, all of your data storage in your um enterprise environment. We wanna create developer self service for you
15:22
so that you can create policies for how your data in your organization is managed and then allow your developers to get access to those managed storage classes you're in control. But so are they uh and it really, it builds a beautiful relationship. Um We want a seamless data platform. Why have multiple operating systems when one will do? And then finally, this will allow you to save
15:50
money because instead of having some arrays that are full and some that are empty, but you can't share the the storage, we can now treat the storage in your enterprise as a utility, all of it operating together satisfying your data storage needs. So this is our data set uh storage platform. Let's see how it works against these new challenges that we talked about.
16:15
So starting with A I with A I, this provides performance at scale. The world's largest G uh GP U clouds, the world's largest GP U supercomputer. The meta research super cluster relies on pure flash blade to deliver optimal performance for their GP us running at 100% even better. We can provide a performance across all of our tiers of storage.
16:41
If we replace your spinning disks with flash for roughly the same uh price at entry, it's five times more performant. You have extra performance for things like access to A uh to A I analytics. And we are building for you a broad set and you'll hear this more at the event. A broad selection of NVIDIA um validated designs uh for rag and for inference and many
17:07
of them very much focused on vertical uh business, uh businesses such as health care, such as pharmaceuticals, banking, manufacturing, et cetera. Ok. Let's look at cyber security. Pure was the first offer, immutable snapshots and rapid recovery from cyber attack. Now, as you'll be here later, we're gonna be offering life cycle support and emergency
17:32
support for your clean room environments when you have law enforcement also involved uh in any cybersecurity attack. All right, let's take a look at um at application modernization. As you are looking at moving on from virtual machines to containers, we are able to provide you the ability to do um uh to be able to make that motion,
17:59
whether it's moving from one type of VM uh environment to another, whether it's about moving from a on prem VM environment to a cloud VM uh environment or whether it's moving from V MS uh to containers. We have the technology to make it easy for you to do so. And then finally making hybrid cloud real, we have purity working uh in Aws and uh uh and in Azure uh
18:26
that allow you to move your and operate your uh VM based uh workloads in both environments with exactly the same services exactly the same API S as we have on Prem that makes the movement of not just your data, but your application easier as well. And oh by the way, we save you a ton of money when we make that uh when we make that move.
18:52
So that's the basic uh platform. It's what makes the platform a platform, what do you think?
