 They're sometimes missed But they never miss a thing Obsessive meticulous Ingenious human sponges reload They don't sleep you need sleep Imagination running wild Delerating faster until they hit a wall The app data gap. We don't like ceilings. We understand them because we are them Introducing the nimble storage predictive flash platform and the predictive analytics of info site With flash performance to run your application seamlessly Delivering maximum data velocity So you can run your business Your research your city Please welcome Janet Matsuda nimble storage CMO. Good afternoon I wanted to thank all of you who are joining us both on the wikibon broadcast and here in the room in San Francisco I wanted to especially recognize the journalists who've joined us from around the world from Australia From France Germany the UK and Italy. So thank you The man I'm going to introduce now has spent two decades in the storage industry He joined nimble early on as an advisor on the board and Then shortly after we introduced our first product. He came on board as CEO. I'd like to introduce Suresh Vasudevan. Thank you Janet Let me also start off by Welcoming all of you on the broadcast welcoming all of you here. Thank you very much for joining us today Today is a momentous milestone in our journey as a company Our journey began In 2008 when we were founded as a company We got started with the belief That there were some technology disruptions that were so fundamental that we could append the storage industry by leveraging those disruptions That incumbents like EMC, NetApp, HP would not be able to leverage those disruptions And would not be able to adapt to the changes that were coming in our industry as I think back at the catalyst The first catalyst is flash We believed that flash would allow us to accelerate our customers applications by a factor of five to ten times While leveraging the right architecture. We could also lower the total cost of storage ownership the second belief we had was in the power of predictive analytics cloud-based predictive analytics our belief was that we could watch over our customers infrastructure from the cloud and Radically simplify operations improve the productivity of our customers IT teams With that as our founding beliefs we came to market in August of 2010 It's almost been five and a half years that we've been in the marketplace and as I look back on the last five and a half years Perhaps the aspect that we are more proud of as a company than any other aspect is the pace at which we've grown our customer base Today we have seven and a half thousand customers deploying us in over 50 countries around the globe They come from a variety of different industries and they deploy us for a wide range of workloads Within our customer base are thousands of mid-sized enterprises Hundreds of large global enterprises Hundreds of cloud service providers that leverage our product as the underlying building block to deliver cloud services Another measure of the impact that we've had on the industry is When Gartner named us last year as a leader in the magic quadrant in the general purpose to rich magic quadrant And as you look at the leaders quadrant We are the only company Founded in the last 20 years to make it into the leaders quadrant a Measure of the impact we are having on our customers a measure of the impact We are having in terms of innovation that we've brought to bear in the marketplace and today continues that Trend towards broadening our technology innovation that we are introducing in the marketplace this morning We introduced groundbreaking innovation in the form of the industry's first Predictive all-flash array a key component of our predictive flash platform Now before we dive into specific aspects of our platform I thought be useful to step back and take a look at how we view the customer problem that we are solving as individuals each and every one of us interacts with technology on a daily basis and We've now come to expect that information is available to us instantly a great example Is when I get into my car in the morning type in a Google Maps address My expectation is that I would have access to the address immediately to the route immediately When something slows down the delivery of data when something Disrupts the application we all tend to get frustrated. We believe that our time is wasted that our productivity is impacted Well organizations are not that different Modern organizations rely on hundreds of applications that power all of the core business processes Inventory management financial reporting how you support customers how you build products all of the core business processes in any organization rely on Applications and those applications in turn rely on timely rapid access to underlying data So when something comes in the way of delivering data to the application what you're really impacting our business outcomes this gap between the application and The speed of data delivery is what we refer to as the app data gap So the interesting question then is what causes the app data gap Now at nimble across over seven and a half thousand customers as we've looked at what causes applications to be disrupted It turns out that storage which Lies at the bottom of the infrastructure stack is indeed a key component Less than about 45 percent or so of all the problems that disrupt applications stem from issues in the storage layer Not enough performance software issues hardware problems But as it turns out when you look at what disrupts applications over half the problems are from the rest of the infrastructure stack Not having compatibility between compute networking storage not having adherence to best practices inadequate compute resources and so on infrastructure complexity is as critical if not more critical as a barrier to delivering Predictable well-performing applications as storage itself so the question then becomes how do we address this app data gap and As we sought to address that problem. I want to go back to the beliefs that we started with as a company The first enabler to addressing the app data gap is really flash Flash is hundred times faster than disk drives and a well architected storage systems that system that leverages flash Is able to accelerate applications by a factor of five to ten X while still lowering the cost of deploying storage But as we've just seen having fast storage alone is not enough Flash alone is not enough in ensuring that you have applications that perform Well businesses that perform well a key part of the problem and a key part of the solution is how we address Infrastructure disruptions and that's really where the second belief we had comes into play If we could monitor our customer deployed infrastructure from the cloud on a real-time basis gather information about the health of that infrastructure and use predictive analytics to Predict what will break down and prevent those break downs from occurring then you address the second part of the app data gap and That in a nutshell was the belief system we started with to address the app data gap You need to leverage flash and you need to leverage predictive analytics and those beliefs What we translated into our predictive flash platform The predictive flash platform that we've brought to bear has three key components The first component is infocyte This is our cloud-based predictive analytics software that watches over our customer infrastructure and predicts and prevents issues The second component is what we are calling our unified flash fabric a Single software architecture that spans our adaptive flash arrays and all flash arrays and provides a single Consolidation platform such that you can accelerate every application in the enterprise while lowering the cost of ownership of storage That's the second component of our predictive flash platform The third component is what we refer to as timeless storage Conceptually what timeless storage accomplishes is that when our customers make investment in our products? We want to ensure that we're delivering investment protection to our customers and that's really the aim of timeless storage Now when I translate The predictive flash platform into what it means for our customers. What do our customers experience when they deploy our predictive flash platform? There are a few benefits that stand out as Distinguished in when they deploy our platform the first is in terms of non-stop availability Across our seven and a half thousand customers. We have a demonstrated actual measured availability of ninety nine point nine nine nine seven That's eight seconds of downtime per system per month So the first benefit is non-stop availability measured in terms of uptime measured in terms of data protection Measured in terms of data security and that's really the first thing that our customers see The second benefit is absolute performance every application Speeds up by a factor of five to ten X experiences less than a millisecond latency Across the entire enterprise and that's the second big benefit that we bring to bear The third benefit is cloud-like agility Freeing up time from our customers IT teams so that they don't have to manage infrastructure They can spend time optimizing infrastructure for the needs of the business Enhancing their productivity and that's the third benefit that our customers experience and lastly even as we deliver non-stop Availability even as we deliver absolute performance even as we deliver agility We do so while lowering the cost of storage by a third to two third Compared not just to legacy storage platforms, but also compared to more modern storage systems So that in a nutshell is what our customers experience when they deploy our predictive flash platform Now over the next several minutes, we'll walk through every element of the predictive flash platform Rod bag the person that built infocyte will come and talk about infocyte Dan Leary our head of products and solutions will talk about our unified flash fabric Let me first briefly touch on the concept of timeless storage So as I said earlier timeless storage is all about ensuring that our customer satisfaction is unmatched That we provide investment protection for our customers and it's made up of multiple elements the first and perhaps most significant element of timeless storage is a maniacal focus on ensuring that customers are successful when they deploy our products One part of making that happen is making product claims that are transparent where we deliver more than what our customers expect From the claims that we make about our product The second part of that is really about ensuring that we provide proactive support that is ahead of anyone else in our industry Now it's easy to make these claims Perhaps the best proof point of how well we live these claims is in the form of our net promoter score Many of you already know the net promoter score is often a gold standard for measuring how how we deliver customer satisfaction How a vendor delivers customer satisfaction in our industry and in the net promoter score survey the inaugural survey that we did over the last couple of months our net promoter score stands at an astounding 85 and Just to put that in context companies like EMC companies like net app IBM HP and others have net promoter score in the 30s and 40s a true testament to how well we ensure that our customers are successful The second element of timeless storage is in the form of guarantees that we make to our customers The fact that all of our software is bundled in as product as part of the product and we don't nickel and dime our customers on software licenses We don't escalate the price of support after year three and years four and five because we want to make sure that our customers Don't have to go through forklift upgrades We've also introduced a support option whereby a customer is entitled to a controller upgrade after three years Again, so that they are not forced into a forklift upgrade after the initial support term of three years is done So guarantees that make sure that our customer's investment in our products is protected is the second element of timeless storage and The third element is really in terms of flexibility in how our customers deploy our products Like anyone else in our industry our customers often buy our products spend capex and deploy products in the data center What we offer that's also unique is that they can choose to Deploy our storage systems as a service where we monitor their actual usage on a monthly basis and Build them for that what they actually use whether that usage goes up or goes down So deploying storage on demand not as capex but as op-ex is another element of timeless storage So thus far what I've talked about is our predictive flash platform being made up of three Components infocyte unified flash fabric and timeless storage. I've briefly touched on what timeless storage means for us. I Want to go back to the announcement that we made this morning We as I mentioned we announced the industry's first predictive all-flash array a Market-leading we believe an all-flash array that leapfrogs over every other all-flash array in the marketplace And over the course of the next several minutes you'll hear more about the details of the all-flash array at a summary level what we've brought to bear is a system that Stands out for its scalability Scaling up to eight petabytes in a single system delivering over a million IOPS Less than a millisecond latency for hundreds or even thousands of apps That's the first aspect in which our all-flash array is very distinct The second aspect is really around the cost of deploying all-flash arrays Our all-flash arrays typically are about a third to two-thirds lower than market-leading all-flash arrays And that's through a combination of innovative innovative software that powers our all-flash arrays And the last aspect that I'd call out as very distinctive about the all-flash array is is what I mentioned earlier We already have demonstrated availability of ninety nine point nine nine nine seven and our availability continues to improve every quarter The same nimble operating system that powers our adaptive flash array also powers our all-flash arrays And so the third benefit is really in terms of availability Now before we dive deeper into infocyte and dive deeper into our unified flash fabric and the all-flash array At this point I'd like to pause Invite Varun Mehta our founder employee number one and head of engineering to come in and have you see physically What the all-flash array looks like Varun? So On behalf of all the employees at nimble, I'm very honored to present to you the new leader in all-flash arrays And turn it on to thank you Varun Thank you It's truly an exciting day for all of us at nimble for our partners as well as our customers and This announcement of the first predictive all-flash array and the predictive flash platform marks a milestone in this Transition that's happening in the storage industry. It's truly disrupting the storage industry and So with that our next speaker. I would like to invite up. He was one of the first 20 employees at nimble and he had the vision that we could transform the customer support experience by Combining cloud connectivity with big data analytics He's not only the father of Info site, but he wrote a lot of the initial code I'm pleased to introduce rod bag our vice president of analytics and support All right, thank you very much for the introduction and welcome to everybody for being here So I wanted to just you know step back So we talked already about the two big elements of this being the new flash Platform all flash platform and the predictive analytics that sit on top of that So I want to cover Info site right now in just a little bit more detail So since our inception seven years ago or so with nimble OS we have actually embedded Thousands and thousands of comprehensive Sensors within the operating system to actually collect send back massive amounts of health information For every array that's installed throughout our install base all of that comes to us in real-time and goes through analytic Analysis literally when we get that information and to provide information back to our customers So they don't have to look at all of that data. They can look at what we're predicting for them and making recommendations Just to put it in context If you take a look at all of that data that we get back It's somewhere in the order of about a hundred billion Pieces of data every four hours that we're collecting that's more than the number of stars that there are in the galaxy Oops I'm sorry. Can you go back one? Thank you. So it's as Suresh pointed out The problem is not just with storage. In fact when you talk about the app data gap Less than half of the issues that we see from our infocyte data is actually related to the storage element the majority of it is Elements outside of the storage infrastructure, but other pieces of that infrastructure. So that's the networking. It can be hosts Configuration of virtual machines how all of these things interoperate That's where the majority of the issues come from and if you imagine a IT department Their task is to understand lots of them get their own operational data from all of these elements And their task is to try and understand that make sense of it and do something about it It's very difficult for them to grok all of that information nowadays. It's just too big so what we do from an infocyte perspective is We have taken not just the sensors that we first built into nimble OS But also now Sensors that we have actually built up the entire stack throughout the VM Infrastructure so that we can actually see what's going on across that stack It's not enough anymore to just look at storage as we pointed out in the previous slide Most of the problems are outside of storage. So it's absolutely necessary that we collect that data all the way up the stack So having that cross stack sensors is absolutely essential to taking that and then providing Deep data analytics back to our customers not with just raw data that we present back to them But with actionable predictive analytics that helps them avoid those issues in the first place all of that Loops back through the install-based learnings that we have across thousands and thousands of systems out there in the field to then to then Enhance all of those sensors enhance all the predictive analytics that we do on it to really constantly improve that Scenario for our customers and really avoid that app data gap So when we take a look at all of these sensors that we get and all of the predictive analytics The two most important things to that app data gap is providing data velocity and what we mean by data velocity are two things non-stop availability so preventing issues and preventing downtime in the first place and then providing Absolute performance and so we take a look at all of these predictive analytics that we do and just make sure that we're Analyzing from a performance perspective as well and so that we can predict and make recommendations for our customers to optimize that Now when we talk about a cross stack analysis, I'll give you a quick example of that We've actually been known for a long time now even before we Started calling it this predictive flash platform But we've been known in the industry and are from our customers to really be the experts not just in storage But also up the stack So when we actually have problems that customers call us about Good point good portion of the time greater than half of the time They're not storage related but because we have that information at our disposal We're able to troubleshoot anything up that stack in a very effective fast manner because we have data at our fingertips and the other sort of end of it which I'll talk in a bit is that we actually have no level one or level two Support engineers as a result all of that is automated and with all of that data We can have our level three engineers really do the analysis themselves with all of the data that they have at their disposal The last thing that we do with all of this information is to be able to predict Capacity trending for our customers so that we can notify them well in advance to make plans around Expanding their storage capacity and also performance predictions So we can actually see how those arrays are performing and make recommendations to do hardware upgrades software upgrades To actually improve performance over time before it becomes a problem and of course avoid that app beta gap So the bottom line is these three benefits that you see up on the screen for our customers We already talked about the high availability 99.9997 that's less than 95 seconds per year of unplanned downtime. That's unheard of in the industry We that is actual measured availability across the entire install base across all of our general availability releases out there Next element that is a benefit to our customers is that nine out of ten cases or issues We actually find them first before them and provide a prescriptive answer on how to avoid that issue altogether before it becomes a problem In their environment and then finally I talked a little bit about these level three engineers So I'll give you a you know just one example of that. So we have We had a customer that had an issue in their infrastructure that they knew very well was not a storage related Issue but was something and one of the other vendors in that infrastructure. They actually sent me an email sort of outlining the case progress for that problem when they were dealing with the other vendor and That took six weeks for them to go back and forth with that other vendor Asking them for information asking them to try this and try that took six weeks They finally escalated the case that had been escalated for a couple of days When the customer finally decided to just call nimble even knowing full well, it was not a storage problem We solved that problem for them same day for a problem that took six weeks For the other vendor to not get close to solving and that was just because of the fact that we hire level three talented Engineers that know the entire stack and we're able to do that because of the predictive analytics and that is the real power of infocyte Thank you Thank You Rod Our next presenter was actually employee number six at nimble and he was a critical leader Who helped to grow the business from zero to over 300 million dollars? And so with great pleasure, I'd like to introduce Dan Leary the vice president of products and solutions So I'm very excited to get to speak to everyone today And I'm particularly excited to be able to walk through some of the details of our unified flash fabric and show you how we've done some Things that truly have leapfrogged the best capabilities that are available in the industry today Is this work? Okay, there we go Okay So we talked about this already We're going to focus on what we've done with our new all flash array and I'm going to highlight Three key areas where we've literally leapfrogged the best in the industry around performance and scalability Dramatically reduced total cost of ownership Well at the same time delivering non-stop availability for our customers Before I go into all the details I want to spend just a moment taking a brief look at the evolution of flash within the storage industry For some context on how nimble thought of this approach to the problem We look back about a decade ago We saw flash first enter our market in two primary areas First the legacy companies that largely bolted flash into their existing storage architectures That were designed in a world for disk The same time we saw the advent of the first generation of all flash arrays Companies like Texas memory acquired by IBM as well as violin memory that developed the first dedicated all flash arrays But here because of the high cost of flash They were largely limited to a very small tier of applications demanding extremely high performance Without a dramatic increase in capacity We then saw the second big evolution as companies like nimble came along and by building a ground-up software design To leverage the complementary capabilities of flash and disk could offer a solution with three to five times the efficiency Of what the legacy vendors could deliver The same time we saw the advent of the second generation of all flash arrays from companies like extreme IO acquired by emc in pure storage and Here the focus was really around data reduction and data services With the focus on new techniques like inline compression and deduplication that could reduce the cost of that flash and begin to make it affordable for a broader mix of performance demanding applications Even with these events It's forced customers to make a choice and ultimately have to deploy multiple separate silos of storage Hybrid which is ideal for mainstream apps, but may not provide that guaranteed Predictable performance needed for the most performance demanding apps And at the same time all flash arrays which are ideal for those performance demanding apps But remain too expensive for the majority of mainstream applications And what we're really ushering in with our adaptive with our unified flash fabric and our single consolidation Architecture is the need for customers to no longer need to make that choice and the result is they've now got the ability To democratize and leverage flash across their entire application infrastructure So let's again look at how that Unified flash fabric is put together and it starts with our new lineup of all flash arrays that can deliver Absolute performance for the applications that demand that performance sensitive application But in addition our long-standing lineup of adaptive flash arrays That have the ability to vary the ratio of flash From very small amounts up to a hundred percent flash to give that high performance, but also efficiency Mandated by mainstream applications Because they all share a common nimble OS and a common set of data services Our customers can now manage all of these in a way to solve flash for the all mainstream applications So what I want to do next is focus on each of the three areas that we differentiate on and walk everyone through a little bit Meet more detail in our solution Starting with our ability to deliver something that we call sheer performance and scalability Now one of the innovations that nimble made to the market if we go back a couple of years ago Was the introduction of a novel scaling technique that we call scale to fit What scale to fit does is give our customers the flexibility to scale the deployment of their storage Whether it's performance capacity or both at the lowest incremental cost with full increment with full investment protection And the ability to support always on availability, so there's no disruption and no downtime So for example when this now applies to our all flash arrays It allows our customers to start with a single all flash array And as they grow out the the needs for capacity of that single array They can simply add a series of all flash shelves To grow the perform to grow the capacity of that all flash array with no disruption Beyond that though they have the ability to upgrade the performance Non-disruptively with disruptively within that array by being able to remove the controllers within that array and Replace them with more powerful controllers taking advantage of more Intel cores and more DRAM In effect what this allows our customers to do is start with our entry level all flash system and Upgrade it to our highest end system with full investment protection and no downtime But even more important in all flash What does a customer do if they exceed the performance or capacity limits of a single system? And this is where our scale out clustering technology is so critical allowing our customers to go beyond the performance or capacity envelope of a single array and Allow an application to take advantage of the combined performance and capacity of a cluster of up to four different arrays So allows us to get to very high both performance and capacity scaling limits And let's take a closer look at how this works So what you can see here is a single one of our new all flash arrays? That can now deliver industry leading two petabytes of effective capacity within that single array Or up to 300,000 IOPS and then clustered in a four node scale out cluster Achieve quadruple those limits So here we're looking at over eight petabytes of capacity and over a million IOPS within a single cluster And if we compare that to a couple of the market leading all flash arrays starting with EMC's extreme IO You can see first of all the tremendous advantage that we achieve in capacity scalability Having ten times the scale within a single system of the EMC system 200 terabytes versus two petabytes with nimble and at the same time Achieving double the performance of a single nimble array In fact, it would require an eight node expert cluster From EMC to be able to match nimble's performance with only four nodes and even there the EMC Solution offers less than a quarter of the capacity that nimble can deliver Next I want to compare briefly to another innovator in the all flash space pure storage Again here you can see nimble offers a substantial Scaling advantage in this case five times the effective capacity within a single system 400 terabytes versus the two petabytes of nimble And because pure lacks a scale-out capability, there's no way to be able to scale Beyond the limits of a single system without creating multiple separate silos of storage now Another key component here though and a really unique feature of our unified flash fabric is our ability to deliver this in a seamless way Across the entire application environment a really critical capability is our ability to cluster our all flash And our adaptive flash arrays together within the same scale-out cluster This allows our customers to deploy Applications and workloads on the storage that ideally suits their needs So for example, I could deploy a production workload of Oracle or SAP That needs that predictable sub millisecond latency on all flash arrays While deploying mainstream applications like Splunk or Microsoft Exchange very cost effectively on adaptive arrays Because these say these applications are supported by the same scale-out cluster Customers at the push of a button can transparently move an application From adaptive to all flash arrays or vice versa as their needs evolve So again for example here I can promote a test and development instance of an Oracle database And immediately move it into production onto an all flash array with no downtime and no disruption to operations okay Next I want to spend a few moments talking about how we deliver this dramatic Total cost of ownership advantage and savings for our customers That is a one-third to two-thirds savings over the leading systems in the market today And I'm going to cover now four key architectural advantages that underpin our new all-flash arrays To support this very dramatic total cost of ownership savings And I want to start by talking about memory I'm talking about the DRAM that underpins our all-flash array and every array that's out in the market today And it's not something that's often thought about by customers But it's critical in order to be able to deliver the data services that are so critical in all-flash arrays When you think about things like deduplication requires managing a substantial amount of metadata So these arrays can understand the fingerprints of data being written to those arrays and Determine whether there's a duplicate on the system, so it doesn't store it twice What we've been able to do that's extremely critical here is build a system that is 10 to 30 times more memory efficient than the best available in the industry today and That has a number of very powerful advantages for our customers First it allows us to dramatically reduce the cost of those controllers But even more importantly to allow our systems to be able to address Substantially more flash within a single controller Allowing our customers to be able to deploy more flash To a wider set of workloads with less total controllers And the best way to really understand how this advantage plays out is to look at a real-world example Of the new nimble all-flash array compared to our competitors And so for this example what I'd like to do is look at a deployment of two petabytes of effective capacity And what you can see first on the left of my slide is a solution with a single one of our high-end What we call nimble af 9000 arrays along with two disc shelves That supports two petabytes of effective capacity in just 12 rack units of space And if we compare that to pure storage They would actually have to deploy five of their high-end m70 controllers Along with 20 expansion shelves requiring more than four times the rack space To be able to support the same amount of effective capacity and Remember because pure doesn't offer scale out. That's actually five separate systems that have to be managed by customers Now if we next compare that with emc extreme. I owe You see even even more dramatic difference emc requires 66 rack units and 12 separate experts in order to be able to deliver a comparable amount of capacity and If we finally look the solid fire the recent acquisition by net app It's an even more dramatic Difference 82 separate controllers Requiring 82 rack units to deliver the same capacity So that's I think a real visual reminder is to the critical importance of having built an architecture That's unconstrained by memory in order to allow flash to be deployed in a broad way across the enterprise Okay, the second key area I want to explore is our ability to be optimized and designed For the use of 3d NAND this is I think very well understood in the industry It's been rapidly evolving to leverage 3d NAND because that's a critical way to increase density and reduce the cost of flash However, many of the new forms of 3d NAND support lower endurance as a cost to get this higher density and lower cost And one of the foundational elements of our file system that goes back to the earliest days of our initial system is Our ability to convert random rights or random IOs and coalesce them into a single sequential Stream before they're written down to the underlying flash media This allows us to essentially significantly reduce the amount of right amplification to the flash Which in turn improves performance and Allows us to guarantee a seven-year lifespan for those SSDs in customer environments We've also done a number of things like innovative use of integrated hot spares They can improve the usable capacity of our systems by at least 20 percent over the leading competitors So the result is lower cost and more flash with more efficiency Okay, next I want to speak for a moment about data reduction and data reduction is a particularly critical area within all flash Because in spite of the fact that flash has gotten significantly less expensive in the past several years It remains eight to ten times More expensive per gigabyte than the cheapest densest forms of disk And so a fundamental underpinning of our all-flash array and another key innovation was our use of a unique variable block deduplication variable block compression and Zero pattern elimination that provide a very highly efficient But also highly effective form that can typically reduce data by 5x before it's written down to the array and With coupled with capabilities like our thin provisioning and zero copy clones our customers have the ability to save even more capacity Okay Finally, I want to speak to data protection We've been speaking a lot about primary storage, but an equally critical area is data protection And one of the other unique capabilities that we bring to bear with our unified flash fabric is the ability to replicate data from a primary all-flash array to a very highly cost optimized adaptive array So our customers are able to take very frequent application consistent snapshots on their primary data center and very efficiently replicate those They're a feature that we call smart replicate to a cost optimized hybrid array located at their dr site In doing that they're able to achieve very stringent recovery point objectives But because of the efficiency and low cost of that hybrid array it becomes now cost-effective to store weeks Or even months worth of snapshot based backups at the remote site And we've done a lot of work to partner with leading software companies to be able to provide a complete data protection solution for our customers here Okay Finally, I want to touch just for a few moments on what we've done to deliver absolute resiliency for our customers And you've already heard a lot today About the power that we brought to bring a very remarkable amount of availability more than five nines of availability And again, that's underpinned by both our redundant hardware That offers no single point of failure our very robust software architecture And of course our info site platform that gives our customers less than eight seconds of downtime per array per month We've also innovated deep within the software stack for our new all-flash arrays in Delivering a unique capability that we call triple-plus parity raid So unlike competitive all-flash arrays that can tolerate the loss of one or at most two SSDs Our solution allows our arrays to survive three failed SSDs and continue to operate without interruption But beyond that the plus and triple plus parity raid is our unique ability to offer an Intra drive parity Across the different SSDs that allows them to recover from sector errors even with three drives failed The result of that is the highest availability and the highest degree of protection available in the industry Now I've already spoken to the integrated data protection and our app consistent snapshots and replication Made even more powerful with the ability to support that replication between our all-flash and our adaptive arrays And then finally a feature that we added last year for our adaptive arrays that we're bringing forward to the all-flash Market, which is ideal for our most secure customers is a software based what we call smart secure encryption capability This offers appulate application granular encryption and data shredding Including the ability for our customers to replicate from a primary data center to a dr site while being fully replicated and to end Next I want to spend a moment talking about the specific lineup of new flash arrays that we've introduced And we're really excited here to be able to offer a very complete lineup of AFA's Starting with our AF 3000 and our AF 5000 that offer a very cost-effective way for mid-sized Enterprises or smaller deployments who want to take advantage of the deterministic low latency of flash to do so very cost-effectively All the way up to our leading af 7000 and af 9000 that offer absolute performance And the ability to consolidate hundreds to thousands of workloads onto a single system So hopefully you've been able to see here why we're so excited about this announcement today And why we think we've done some things to fundamentally disrupt the market bringing a huge improvement in performance and scalability Dramatic reduction in total cost of ownership all well delivering non-stop availability for our customers. Thank you very much All right. Thanks Dan all right, so We are privileged to have today with us an analyst Actually, I did have two more slides that I think you skip past if you can go back They go away predicted flash platform the one after that. No You go forward to All right So Dan did a great job talking about the unified flash fabric and talking about our all flash array So just wanted to remind you that wrapping around that is the predictive flash platform And so for our customers what that means is non-stop availability Absolute performance and cloud like agility and everyone wants that at a lower price, right? Absolutely, so that's what we're able to deliver today and For nimble, this is a great opportunity Because today we're really at an inflection point in the storage industry and there is an opportunity For nimble to be this the leader in this new era of predictive flash And what it means for our customers is that they're able to deliver data velocity to their users to accelerate their business processes and To gain competitive advantage for their organization So we're very excited and thank you for joining us today to hear these announcements