00:01
Mhm. Hello, everyone, and welcome to this session on Evergreen Storage and Evergreen Flex. My name is and ignorant. I am vice president of product management and pricing strategy. Today I am joined by Ripple Bhatnagar.
00:23
Ripple leads the Monetisation strategy for pure. We will spend the next 30 minutes. Talking about are evergreen architecture and how we are extending it across the portfolio with the introduction of Evergreen Flex. I have had numerous conversations with customers where they talk about having to replace storage arrays every five years or so because they are a has gone end of sale.
00:52
This is definitely not the case with pure storage areas. With the introduction of our evergreen storage architecture over a decade ago, they got rid of this forklift upgrade cycle. In fact, you're built a storage system that gets better over time. Our hardware and software are built from the ground up to be upgraded non destructively,
01:15
with no downtime or data migration. This is not something that can be retrofitted into outdated architectures. It needs to be purpose built into the product. Our customers can modernise their pure storage arrays over multiple generations while keeping the investments protected.
01:35
Numerous pure customers like Sierra Nevada and Mata site have been benefiting from this since the very beginning. This is also validated by the fact that over 97% of our arrays purchased six years or more ago are still in production. Are evergreen Architecture has given us not only the best products in the industry, but also given us the flexibility to offer our customers unique and innovative procurement
02:07
models? All the benefits of our evergreen architecture are available to our customers through our evergreen subscriptions. This gives our customers a subscription to innovation, which keeps their pure are is modern and constantly evolving. This is still unmatched in the industry. Even a decade later,
02:30
we overlay this with pure one, our AI ops and cloud based monitoring solution, and made our storage areas available as a service with cells in the form of pure as a service. With the introduction of Evergreen Flex, we are now extending our evergreen architecture to our customers fleet of pure areas. We will talk about this in later slides.
02:58
Our procurement options give our customers flexibility in how they subscribe to pure evergreen architecture. So along with the introduction of Evergreen Flex we are bringing all these procurement models under the evergreen umbrella. As part of this change, we are renaming our current offerings.
03:20
Evergreen Forever is the new name for our Evergreen Gold subscription. Evergreen Forever continues to be available with our Capex ownership model. Evergreen one is the new name for Pure. As a service, Evergreen One is a pure, own and manage service for customers that want a cloud like experience for their data storage.
03:47
Evergreen Flex is the new addition to our evergreen family, where we extend the evergreen architecture to a customer's fleet of pure storage arrays. Customers' own the hardware but consume the software and services as a subscription. Let's take a deeper look at Evergreen Flex. With our Capex offerings, we have an all inclusive products.
04:13
Q, which combines a hardware and software trophy entitlement, is perpetual in nature and goes with the product offering. We then have the term based Evergreen subscription, which gives our customers a subscription to innovation, which keeps their products modern with non disruptive hardware and software upgrades.
04:37
Around four years ago, we made a storage or is available to the market as a service based on a usage based subscription with Evergreen one Pure only manages the hardware, and our customers get a true OPEC's cloud operating model and service with zealous. We have been very successful with both these models in the market. However, through a journey with these two models, we did identify a few gaps.
05:12
When a customer buys a new array, it is not easy for them to predict how that arias utilisation will grow over time. Unpredictability in demand can result in over utilisation of one array in the fleet, while another array remains slightly used. This is a pain point for customers, and they don't like seeing their large investment
05:35
sitting under youth line. This was one issue we wanted to address with Evergreen Flex. Then there are a set of customers who favour the paper use economics offered by Evergreen one, but must own the hardware. There are several reasons for this. In some cases,
05:55
there are regulations that require the customer to own the hardware. Some customers have tight change control processes, which prevent the use of just in time edition of capacity. Others might require complete control on which hardware and how much capacity is deployed. There are some customers who like the low upfront costs of offered by Evergreen one but
06:21
operational ising are full as a service is a challenge. And for some customers, there are financial aspects which require them to keep assets on their books. To better meet the needs of these customers, we are introducing Evergreen Flex in the traditional appliance model. The hardware and software are bundled together in Evergreen Flex.
06:48
We separated the software from the hardware. We took the software piece and combined it with the Evergreen subscription and build a usage based service subscription. This very much looks like the Evergreen one subscription, but now customers can buy and own the hardware and pay for software and the evergreen subscription in a pay per use model.
07:13
The customer now has full control on which hardware thereby how much capacity they want on site and when they want to upgrade or add more capacity. Mhm with the separation of software from the hardware, the data packs now had the same price across all models. So now with Evergreen Flex, we are enabling data packs to be physically moved between
07:39
arrays of the same flash family. We are basically extending evergreen to the portfolio with a lot of vendors. Capacity additions to a production array are just not possible. With pure, a customer can buy additional capacity and non disruptive, Lee added to the array.
08:01
Evergreen Flex extends this capability and allows customers to move data packs within their fleet from an underutilised array to where they are needed. The most on this slide are screenshots from a pure Wanganui. The cherry on top is 89% utilised, while the one at the bottom is only at 9% utilisation. In this example, the customer can move.
08:30
The 256 terabyte data pack from the underutilised are a to the area on top, where the capacity is required. This is unique in the industry, so now, instead of moving an application when it runs out of storage, we're enabling capacity to be moved to the application from an underutilised already the flex in Evergreen Flex and for the flexibility that this model offers.
09:00
As mentioned on the previous slide with Evergreen Flex, we have unlocked the capacity associated with an array and allow moving the storage capacity to where it is needed. With evergreen, there is the inherent flexibility of non disruptive capacity controller and purity upgrades.
09:21
We provide our customers with flexible ownership of the hardware, and they pay for the utilisation of the asset. The next big thing we are offering is a fleet level subscription with site level aggregation of reserve capacity. Now, an aria in a fleet does not go into on demand building until the aggregate reserve of
09:44
the fleet is used up. There is the flexibility of having a lower reserve commit and consuming additional capacity in an on demand manner to deal with demand spikes. Not to walk you through details on how Evergreen Flex works. I'm going to hand it over to Ripple.
10:03
Thank you. I'm good. And once again, welcome, everybody, um, to the launch of Evergreen Flax. My name is viable, but ***, I'm a senior director for Monetisation Strategy and Deal operations. Uh, pure. So I'm gonna talk about how how this works.
10:23
Everything in flex is essentially a buy and subscribe motion, like I'm gonna explained earlier. And you are essentially buying a site level subscription, not an array by a rare appliance by appliance subscription were essentially buying a site level subscription for choice for your own hardware essentially so you would start with.
10:44
I didn't define sites where you want to deploy pure. You will buy hardware of choice as Capex, and this hardware of choice can be across our portfolio of flattery X flash. We see flashing exile and flash blade. Now what you will notice is that the cost of this upfront purchase for the hardware is lowered by about 70 to 80% depending on different platforms.
11:14
And that's primarily because the cost of this upfront purchase does not include software. So this is essentially just the hardware that you're buying. Of course, with the hardware. Then you do have to subscribe to flex subscription, which then provides you software and services entitlement, which allows you to use that hardware capacity that you have deployed on these sites.
11:40
The minimum dome for the subscription is 36 months, and you essentially get a dollar per effective TV bytes per month as a usage rate. Now this dollar, part to be by per month, is different for different models or controller models that we set. So an X 50 flash array, X vs and si 60 flash arrays C will essentially
12:09
have different usage rates, but you have very consistent dollar per TV pipe per month through the life of this contract and this subscription. The other thing is you have to start with a minimum 30% of the total effective capacity that you have purchased under the subscription as your site reserves and I'll talk about reserve and on demand model. But essentially, a reserve is a commitment to
12:39
use that much capacity for the term of the subscription and on demand usages any usage beyond that result. So it's really 30% and that 30% is typically driven by data that we have seen across are installed base where the utilisation typically trends around 30% in the first year of acquiring hardware, and this is again at a whole site level,
13:06
not individual arrays and appliances. And then you obviously have the record billing, which can be set up monthly, quarterly or annual whatever works best for your business and your budgeting cycles. Now, once we once we map these underlying hardware with the subscription that you have purchased pure is meeting that usage on a daily basis.
13:33
So we do take an average usage for every given day and we would aggregate the usage on a per site for each controller classes. So if you have five X fifties or or four x seventies, we are essentially aggregating the usage on a poor model apart controller class, so that if you have any utilisation imbalances between different arrays that gets neutralised, one array is running 40% utilised.
14:04
The other array is running as 20%. Utilised. The site level utilisation is still within that 30% so you don't incur any additional charges and then you will have access to a customer success manager who essentially will help you be the single point of contact to help you manage the subscription and optimise your experience. Now the on demand and reserved
14:33
construct. It's very unique that we had launched the evergreen one 40 years ago, and the way this this is meant to be providing a lot of flexibility to our customers where you should start with the initial site. Result commit, which essentially is the minimal capacity that you intend to use in a very short duration within safe first 3 to 6 months and that's all you need to commit for you should
15:02
not be looking at a forecasted usage over time. Now you can. You have to still buy that physical capacity in this case. Right? So that pink box that you're seeing, that's the total capacity that you have to purchase physical capacity that you have to purchase and make it available at your site.
15:23
But the software and services you can reserve starting at 30% and let the data come, uh, and be written on these appliances over time in an on demand fashion that's you're essentially are orange curve, and then you have a choice to up your reserve quarter. Minister. The original contract length. You can up your reserve at any point in time because by increasing your reserve,
15:52
pure is going to offer you a discounted rate for your own demand. Consumption typically reserved commit would be about 30% cheaper than the unit prize for on demand consumption, but you have full flexibility. If you have more seasonal workloads, you're better off just leveraging the arm demand usage curve, or, if you have data that's gonna sit there, then you do want to up your reserve and
16:19
optimise your cost for the remaining term of the contract at the end of the term. At the time of renewal, you have full flexibility. You can go back all the way to 30% minimum, or you can adjust as you did necessary. Now it's very important to again highlight that this data that the orange curve that we're looking at this is the data written on the arrays.
16:44
So this is essentially host written data, and this is not the provision or allocated storage. So you could see the benefit in that because you can go ahead when you deploy all this physical capacity, you can actually allocate and provisions a lot of storage of thin provision, a lot of storage out to your host. Well, what you're really what we are really meeting on.
17:08
And what you're paying for is essentially, when data actually gets ingested, the host right? Or if you're creating any kind of snapshots and the unique data from that snapshot that also impacts this orange curve, that's what we are really billing the customers on. It's very important to separate out the adoption of your storage platform,
17:31
where your provisioning, all that storage versus you're paying majority of that payment from a cost standpoint is essentially meant that data actually comes on the area that essentially drives this whole consumption economics in more of a Catholics world, right? This is something which we have had and various other cloud service providers have been providing in dumb,
17:58
soft, painful what you use. But what we're providing is essentially a much more flexible Capex kind of an option where you're doing an up front porches and then paying for software and services in a more consumption manner. So what you see on this graph just comparing different options that you will have now mhm.
18:20
You know, the ground from the left is showing. The green bar is essentially the traditional Catholics model where you purchase upfront. It's high up front acquisition cost. And then you have to only worry about after three years when you're doing a renewal for evergreen forever. Peace now, typically, what we have seen is our customers grow at about 30%. 1st year utilisation curve 50% by end
18:47
of second year and 65% by end of third year and then plateaus there. Now, what you will see here is that orange bar is essentially the flex subscription, and what it's allowing you to do is actually purchase your hardware much more efficiently upfront and then set the reserve as 30% of the capacity and grow that reserve or consume on demand over time.
19:17
And that's your orange curve, which is showing you the cumulative cost. So break even will be somewhere around three or four years if you're following. If your site level utilisation is following the trend that that I'm highlighting there as quick growth Mhm, what you see is that the blue curve that's essentially appeared as a service, which is actually more economical and lower cost of country.
19:44
If you if you go that route because you don't have to buy any hardware, pure buys the hardware. It applies the hardware and provides you a service. Now, over time. Has a pure as a service are evergreen. One is going to most likely come out more expensive than ever green flecks, primarily because there is more SL management and higher
20:07
level of service that we are providing there. But the upfront cost and in a growing environment where you have to do constantly physical hardware purchases ever gain. One appeared as a service is actually still the most recommended option now. What also happens is there are two important data points. If you look at number one there, I said,
20:33
the cost is is much lower than near one. That's about 60% lower than a traditional Capex. That's a huge savings. That's really what the benefit of flexes that you're essentially able to get pure storage technology control, how much capacity and what model you want on site and still reduce the cost and
20:58
really dire cause to the actual consumption of that storage. Now what that does is if your utilisation is lagging and you have a much more slower growth than what you anticipated, then that obviously translates into a much longer break even point and hence savings. In any case, you're most likely your consumption and your cost.
21:24
Directors are essentially aligned directly, and that gives you a flexibility and how you want to procure and acquire pure storage technology. The left hand side kind of shows you the cloud operating experience, and the right hand side is more around financial drivers fitting into your budget, giving you flexibility around that.
21:50
If you focus on the left hand side if really the long term. If you If your utilisation is gonna be pretty quick and you want to basically maintain a lower depreciation expense over a long time. Evergreen forever and an upfront hire purchase is perhaps the best model. It's all up front cost, and but you can with Evergreen forever.
22:16
You can have really long depreciation cycles because this equipment never end of lives, so there's a lot of useful value out of that equipment gives you really optimal unit economics over time, if your utilisation is is much more predictable. But on the other side, if your stakeholders are looking for a lot of agility and cloud operating experience, then you'll ever green one is the most optimal offering.
22:43
Start low and really deliver a cloud operating experience to your stakeholders and optimise your cost, all right in in close alignment with how you're consumption, actually trends so that essentially becomes aware of a green one. And this is all OPEC's solution. So that's another consideration from a budget fit standpoint.
23:08
But now, right in the middle, you have this evergreen flex offering, but still primarily leverages your Capex budget so you don't necessarily have to go right away to optimise your or or rearrange your budgetary considerations, but you still have the ability to now differ as much as 60% of the year. One cost to future years, Um, by actually starting low in terms of
23:38
your reserve commitment, you know, compared to the utilisation of the entire asset that you're purchasing. So it does help your defer your costs and get an optimised return on invested capital very, very effectively. Now it's important to also highlight that one of the key advantage and benefit that we are bringing to market of a green flexes that at any point in time,
24:05
when you are ready to move to Evergreen one and move to a completely cloud operating experience, we will be able to support that transition. And this really helps customers who want to ease into a paper, use economics on premises and then plan for moving into the cloud. It's almost like a stepping stepping stone to get to a complete cloud operating model,
24:29
so there's a lot of flexibility in what fits best for your environment, your your operational and finish up a setup, but then really gives you a transition path to go into Evergreen one, which which we believe is they should be the non star for every idea organisation, and that kind of just, you know, to help you with that hard process.
24:55
Here are some good pointers on you know how to evaluate a good operation and specifically with Flex. If you're struggling with low utilisation, there is a lot of customers by a lot of capacity from us physical capacity from us and in all fairness, all that capacity gets provisions very quickly. We see the strength.
25:18
But that provision capacity does not necessarily translate into an actual data coming in from the host, and that typically is the lab. Now you see a lot of underutilised assets sitting at each of your installed locations so you could avoid high upfront Capex pan and and and really optimise your economics with flex.
25:44
So that's a huge, huge benefit to really track your expense with your, um, utilisation train. But at the same time, have the physical capacity on site two provisions as much storage as you need to your to your stakeholders and host. That actually becomes really powerful value proposition, for if you're a service provider.
26:09
In addition to being able to optimise your cost based on your consumption, what what flex also provides, given that this is primarily a Capex offering and Capex budget is going to be used to buy flags, this is not an optics offering. There's a minimal impact impact on EBITDA and that too many customers, especially managed service providers, is an important consideration.
26:37
So if you don't want to impact your ability to, but you still want to optimise your cash flows and your margins, there's no better model than flex, and you can really align your cost to your revenue. And in fact, we think that you can maximise your margins by driving your revenue based on the provisions capacity while your cost is based on the consumed capacity. So there's a lot of favourable
27:05
economics to go in with and really get to a positive free cashflow already early on in the in the second and at the same time, you maintaining full control on the hardware to provide very applications and specific as it is that pure may not be providing as part of the green one. This model also fits in if you are a very asset heavy industries. These are typically construction,
27:32
logistics, hospitality. Um, there are industries where one of the key metric is return on capital invested. We have seen customers who um, prefer more Capex motions primarily because of that. And they are behind in terms of adopting the cloud operating experience of cloud technologies.
27:57
One of the reasons being how the budgets are structured and how the the success metrics are are laid out. But at the same time, you if you are, if you are part of one of those industries, you know, you do know that you are Capex budgets are shrinking. Um, over time, so flex again helps you actually optimise the capital budgets that you have in hand put assets on your books to show return on
28:24
capital invested and similar metrics, but at the same time get cloud operating model efficiencies on premises while setting yourselves up to actually fully migrate or fully transition rather to a cloud operating experience over time. So this is a really good model for you to step in from a traditional purchase upfront model to a more optimised cloud operating model. But it's all driven by the Capex budget
28:57
And in many cases, if we have regulated environment classified data, very high security standards in terms of not being able to operationalise just in time capacity or giving vendors like US access to the data centre to actually provide an as a service model or you can because of those same reasons you can't really put your workloads in public cloud.
29:23
Then again, Flex comes in very handy where you own the hardware, so you retain that hardware. You can establish your own high standards. You can limit the data centre access because then you are responsible for purchasing and making sure the right amount of capacity is deployed as long as you just call home the the appliances called home.
29:44
We can just meet her and build for all the software and services usage, but now you have an alternative to actually fit into your operating model. So just bringing it all into you know what we call this progression of value and flexibility, everything like anger mentioned. It's all evergreen architecture, right? Whether you opt for evergreen forever,
30:11
where you have longer depreciation cycle, most the most effective and efficient way of, of using, storage and paying for storage. If your workloads are very predictable and you are primarily looking at Capex and depreciation expenses, there is no end of life for evergreen forever. So that's that's a wonderful model for certain customers and all the way the same architecture being leveraged in delivering evergreen one
30:43
where everything isolated when outcomes and essentially all our bags. Cloud operating model pure, owns and services the underlying infrastructure, and you're essentially paying for certain zealous and consumption of the storage services. And then you have evergreen flex in the middle, where we are now optimising the value of evergreen architecture for you at fleet level, providing your flexibility to move around the
31:13
physical capacity like hunger, they explained earlier, and really looking at utilisation of the hardware that you're purchasing in the most optimal man so that you can get the most optimised fleet storage platform with all the assets being monitored and build on usage for the majority of their costs. So different models to get you to the evergreen architecture for many,
31:38
many years where you never have to. Then go and do another refresh on another upgrade tive he Our overall stories experience with that Thank you very much for your time for watching this launch video with us. And I hope you have good questions for us and we'll be happy to help you transition to working Flex.
32:03
Thank you very much. Mhm.
