>> Live from Las Vegas. It's theCUBE. Covering AWS re:Invent 2018. Brought to you by Amazon Web Services, Intel and their ecosystem partners. >> And welcome back to our continuing coverage here on theCUBE of AWS re:Invent, we're on day three of three days of wall to wall coverage that we've brought you here from the Sands Expo along with David Vellante, I'm John Walls. Glad you're with us here, we're joined by David Flynn from Hammerspace, and David, good afternoon to you. >> Good afternoon. >> Been quite a year for you, right? >> Yeah. >> This has been something else. Set us up a little bit about where you've been, the journey you're on right now with Hammerspace and maybe for folks at home who aren't familiar, a little bit about what you do. >> So Hammerspace is all about data agility. We believe that data should be like the air you breathe, where you need it, when you need it, without having to think about it. Today, data's managed by copying it between the sundry different types of storage. And that's 'cause we're managing data through the storage system itself. What we want is for data to simply be there, when you need it. So it's all about data agility. >> I need to know more. So let's talk about some of your past endeavors. Fusion-io we watched you grow that company from just an idea. You solved the block storage problem, you solved the performance problems, amazing what you guys did with that company. My understanding is you're focused on file. >> That's right. >> Which is a much larger-- >> Unstructured data in general file and object. >> So a much larger proportion of the data that's out there. >> Yes. >> What's the problem that you guys are going after? >> Well at Fusion-io and this was pre-flash, now flash everybody takes it for granted. When we started it didn't really exist in the data center. And if you're using SAN, most likely it's for performance. And there's a better way to get performance with flash down in the server. Very successful with that. Now the problem is, people want the ease of managablility of having a global name space of file and object name space. And that's what we're tackling now because file is not native in the Cloud. It's kind of an afterthought. And all of these different forms of storage represents silos into which you copy data, from on-prem into cloud, between the different types of storage, from one site to another. This is what we're addressing with virtualizing the data, putting powerful metadata in control of how that data's realized across multiple data centers across the different types of storage, so that you see it as a single piece of data regardless of where it lives. >> Okay so file's not a first class citizen. You're making copies, moving data all over the place. You got copy creep going on. >> It's like cutting off Hydra's head. When you manage data by copying it you're just making more of it and that's because the metadata's down with the data. Every time you make a copy, it's a new piece of data that needs to be managed. >> So talk more about the metadata structure, architecture, what you guys are envisioning? >> Fundamentally, the technology is a separate metadata control plane that is powerful enough to present data as both file and object. And takes that powerful metadata, and puts it in control of where the data is realized, both in terms of what data center it's in, as well as what type of storage it's on, allowing you to tap into the full dynamic range of the performance of server-attached flash, of course Fusion-io, very near and dear to my heart, getting tens of millions of I-ops and tens of gigabytes per second, you can't do that across the network. You have to have the data be very agile, and be able to be promoted into the server. And then be able to manage it all the way to global scale between whole different data centers. So that's the magic of being able to cover the full dynamic range performance to capacity, scale and distance, and have it be that same piece of data that's simply instantiated, where you need it, when you need it, based on the power of the metadata. >> So when you talk about object, you talk about a simplified means of interacting, it's a get-put paradigm right? >> That's right. >> So that's something that you're checking up? >> That's right, ultimately you need to also have random read and write semantics and very high performance, and today, the standard model is you put your data in object storage and then you have your application rewritten to pull it down, store it on some local storage, to work with it and then put it back. And that's great for very large-scale applications, where you can invest the effort to rewrite them. But what about the world where they want the convenience of, the data is simply there, in something that you can mount as a file system or access as object, and it can be at the highest performance of random IO against local flash, all the way to cold in the Cloud where it's cheap. >> I get it so it's like great for Shutterfly 'cause they've got the resources to rewrite the application but for everybody else. >> That's right, and that's why the web scalers pioneered the notion of object storage and we helped them with the local block to get very, very high performance. So that bifurcated world, because the spectrum got stretched so wide that a single size fits all no longer works. So you have to kind of take object on the capacity, distance and scale side, and block, local on the performance side. But what I realized early on, all the way back to Fusion-io is that it is possible to have a shared namespace, both file system and object, that can span that whole spectrum. But to do that you have to provide really powerful metadata as a separate service that has the competency to actually manage the realization of the data across the infrastructure. >> You know David you talk about data agility, so that's what we're all about right? We're all about being agile. Just conceptually today, a lot more data than you've ever had to deal with before. In a lot more places. >> It's a veritable forest. >> With a lot more demands, so just fundamentally, how do you secure that agility. How can you provide that kind of reliability and agility, in that environment, like the challenge for you. >> Oh yeah. Well the challenge really goes back to the fact that the network storage protocols haven't had innovation for like 20 years because of the world of NAS being so dominant by a few players, well one. There really hasn't been a lot of innovation. Y'know NFSv3 three has been around for decades. NFSv4 didn't really happen. It was slower and worse off. At the heart of the storage networking protocols for presenting a file system, it hadn't even been enhanced to be able to communicate across hostile networks. So how are you going to use that at the kind of scale and distance of cloud, right? So what I did, after leaving Fusion-io, was I went and teamed up with the world's top experts. We're talking here about Trent Micklebus, the Linux Kernel author and maintainer of the storage networking stack. And we have spent the last five plus years fixing the fundamental plumbing that makes it possible to bring the shared file semantic into something that becomes cloud native. And that really is two things. One is about the ability to scale, both performance, capacity, in the metadata and in the data. And you couldn't do that before because NAS systems fundamentally have the metadata and data together. Splitting the two allows you to scale them both. So scale is one. Also the ability to secure it over large distances and networks, the ability to operate in an eventually consistent, to work across multiple datacenters. NAS had never made the multi-datacenter leap. Or the securing it across other networks, it just hadn't got there. But that is actually secondary compared to the fact that the world of NAS is very focused on the infrastructure guys and the storage admin. And what you have to do is elevate the discussion to be about the data user and empower them with powerful metadata to do self service. And as a service so that they can completely automate all of the concerns about the infrastructure. 'Cause if there's anything that's cloud, it's being able to delegate and hand off the infrastructure concerns, and you simply can't do that when you're focused at it from a storage administration and data janitorial kind of model. >> So I want to pause for a second and just talk to our audience and just stress how important it is to pay attention to this man. So there's no such thing as a sure thing in business. But there is one sure thing that is if David Flynn's involved you're going to disrupt something so you disrupted Scuzzy, the horrible storage stack. So when you hear things today like NVME and CAPPY and Atomic Rights and storage class memory, you got it all started. Fusion-io. >> That's right. >> And that was your vision that really got that started up. When I used to talk to people about that they would say I'm crazy, and you educated myself and Floyer and now you see it coming to fruition today. So you're taking aim at decades old infrastructure and protocols called NAS, and trying to do the same thing at Cloud scale, which is obviously something you know a lot about. >> That's right. I mean if you think about it. The spectrum of data, goes from performance on the one hand to ease of manageability, distance and scale, cost capacity versus cost performance. And that's inherent to our physical universe because it takes time to propagate information to a distance and to get ease of manageability to encode things very, very tight to get capacity efficiency, takes time, which works against performance. And as technology advances the spectrum only gets wider, and that's why we're stuck to the point of having to bifurcate it, that performance is locally attached flash. And that's what I pioneered with flash in the server in NVME. I told everybody, EMC, SAN, it sucks. If you want performance put flash in the server. Now we're saying if you want ease of use and manageability there's a better way to do that than NAS, and even object storage. It's to separate the metadata as a distinct control plane that is put in charge of managing data through very rich and powerful metadata, and that puts the data owner in control of their data. Not just across different types of storage in the performance capacity spectrum, but also across on-prem and in the Cloud, and across multi-cloud. 'Cause the Cloud after all is just another big storage silo. And given the inertia of data, they've got you by the balls when they've got all the data there. (laughing) I'm sorry, I know I'm at AWS I should be careful what I say. >> Well this is live. >> Yeah, okay so they can't censor us, right. So just like the storage vendors of yesteryear, would charge you an arm and a leg when their arrays were out of service, to get out of your service, because they knew that if you were trying to extend the service life of that, that that's because it was really hard for you to get the data off of it because you had to suffer application downtime and all of that. In the same fashion, when you have your data in the Cloud, the egress costs are so expensive. And so this is all about putting the data owner in control of the data by giving them a rich powerful metadata platform to do that. >> You always want to have strategies that give you flexibility, exit strategies if things don't work out, so that's fascinating. I know we got to wrap, but give us the low-down on the company, the funding, what can you share with us. Go-to-market, et cetera. >> So it's a tightly held company. I was very successful financially. So from that point of view we're... >> Self-funded. >> Self-funded, funded from angels. I made some friends with Fusion-io right? So from that point of view yeah, it's the highest power team you can get. I mean these are great guys, the Linux Kernel maintainer on the storage networking stack. This was a heavy lift because you have to fix the fundamental plumbing in the way storage networking works so that you can, it's like a directories service for data, and then all the management service. This has been a while in the making, but it's that foundational engineering. >> You love heavy lifts. >> I love hard problems. >> I feel like I mis-introduced you, I should have said the great disruptor is what I should have said. >> Well, we'll see. I think disrupting the performance side was a pure play and very easy. Disrupting the ease of use side of the data spectrum, that's the fun one that's actually so transformative because it touches the people that use the data. >> Well best of luck. It was really, I'm excited for ya. >> Thanks for joining us David. Appreciate the time. David Flynn joined up from Hammerspace, and back with more on theCUBE at AWS re:Invent. (upbeat music)

(electronic music) >> Hi everybody, welcome back to Supercloud 2. I'm Dave Vellante with my co-host John Furrier. We're here at our tricked out Palo Alto studio. We're going live wall to wall all day. We're inserting a number of pre-recorded interviews, folks like Walmart. We just heard from Nir Zuk of Palo Alto Networks, and I'm really pleased to welcome in David Flynn. David Flynn, you may know as one of the people behind Fusion-io, completely changed the way in which people think about storing data, accessing data. David Flynn now the founder and CEO of a company called Hammerspace. David, good to see you, thanks for coming on. >> David: Good to see you too. >> And Dr. Nelu Mihai is the CEO and founder of Cloud of Clouds. He's actually built a Supercloud. We're going to get into that. Nelu, thanks for coming on. >> Thank you, Happy New Year. >> Yeah, Happy New Year. So I'm going to start right off with a little debate that's going on in the community if you guys would bring out this slide. So Bob Muglia early today, he gave a definition of Supercloud. He felt like we had to tighten ours up a little bit. He said a Supercloud is a platform, underscoring platform, that provides programmatically consistent services hosted on heterogeneous cloud providers. Now, Nelu, we have this shared doc, and you've been in there. You responded, you said, well, hold on. Supercloud really needs to be an architecture, or else we're going to have this stove pipe of stove pipes, really. And then you went on with more detail, what's the information model? What's the execution model? How are users going to interact with Supercloud? So I start with you, why architecture? The inference is that a platform, the platform provider's responsible for the architecture? Why does that not work in your view? >> No, the, it's a very interesting question. So whenever I think about platform, what's the connotation, you think about monolithic system? Yeah, I mean, I don't know whether it's true or or not, but there is this connotation of of monolithic. On the other hand, if you look at what's a problem right now with HyperClouds, from the customer perspective, they're very complex. There is a heterogeneous world where actually every single one of this HyperClouds has their own architecture. You need rocket scientists to build a cloud applications. Always there is this contradiction between cost and performance. They fight each other. And I'm quoting here a former friend of mine from Bell Labs who work at AWS who used to say "Cloud is cheap as long as you don't use it too much." (group chuckles) So clearly we need something that kind of plays from the principle point of view the role of an operating system, that seats on top of this heterogeneous HyperCloud, and there's nothing wrong by having these proprietary HyperClouds, think about processors, think about operating system and so on, so forth. But in order to build a system that is simple enough, I think we need to go deeper and understand. >> So the argument, the counterargument to that, David, is you'll never get there. You need a proprietary system to get to market sooner, to solve today's problem. Now I don't know where you stand on this platform versus architecture. I haven't asked you, but. >> I think there are aspects of both for sure. I mean it needs to be an architecture in the sense that it's broad based and open and so forth. But you know, platform, you could say as long as people can instantiate it themselves, on their own infrastructure, as long as it's something that can be deployed as, you know, software defined, you don't want the concept of platform being the monolith, you know, combined hardware and software. So it really depends on what you're focused on when you're saying platform, you know, I'd say as long as they software defined thing, to where it can literally run anywhere. I mean, because I really think what we're talking about here is the original concept of cloud computing. The ability to run anything anywhere, without having to care about the physical infrastructure. And what we have today is not that, the cloud today is a big mainframe in the sky, that just happens to be large enough that once you select which region, generally you have enough resources. But, you know, nowadays you don't even necessarily have enough resources in one region. and then you're kind of stuck. So we haven't really gotten to that utility model of computing. And you're also asked to rewrite your application, you know, to abandon the conveniences of high performance file access. You got to rewrite it to use object storage stuff. We have to get away from that. >> Okay, I want to just drill on that, 'cause I think I like that point about, there's not enough availability, but on the developer cloud, the original AWS premise was targeting developers, 'cause at that time, you have to provision a Sun box get a Cisco DSU/CSU, now you get on the cloud. But I think you're giving up the scale question, 'cause I think right now, scale is huge, enterprise grade versus cloud for developers. >> That's Right. >> Because I mean look at, Amazon, Azure, they got compute, they got storage, they got queuing, and some stuff. If you're doing a startup, you throw your app up there, localhost to cloud, no big deal. It's the scale thing that gets me- >> And you can tell by the fact that, in regions that are under high demand, right, like in London or LA, at least with the clients we work with in the median entertainment space, it costs twice as much for the exact same cloud instances that do the exact same amount of work, as somewhere out in rural Canada. So why is it you have such a cost differential, it has to do with that supply and demand, and the fact that the clouds aren't really the ability to run anything anywhere. Even within the same cloud vendor, you're stuck in a specific region. >> And that was never the original promise, right? I mean it was, we turned it into that. But the original promise was get rid of the heavy lifting of IT. >> Not have to run your own, yeah, exactly. >> And then it became, wow, okay I can run anywhere. And then you know, it's like web 2.0. You know people say why Supercloud, you and I talked about this, why do you need a name for Supercloud? It's like web 2.0. >> It's what Cloud was supposed to be. >> It's what cloud was supposed to be, (group laughing and talking) exactly, right. >> Cloud was supposed to be run anything anywhere, or at least that's what we took it as. But you're right, originally it was just, oh don't have to run your own infrastructure, and you can choose somebody else's infrastructure. >> And you did that >> But you're still bound to that. >> Dave: And People said I want more, right? >> But how do we go from here? >> That's, that's actually, that's a very good point, because indeed when the first HyperClouds were designed, were designed really focus on customers. I think Supercloud is an opportunity to design in the right way. Also having in mind the computer science rigor. And we should take advantage of that, because in fact actually, if cloud would've been designed properly from the beginning, probably wouldn't have needed Supercloud. >> David: You wouldn't have to have been asked to rewrite your application. >> That's correct. (group laughs) >> To use REST interfaces to your storage. >> Revisist history is always a good one. But look, cloud is great. I mean your point is cloud is a good thing. Don't hold it back. >> It is a very good thing. >> Let it continue. >> Let it go as as it is. >> Yeah, let that thing continue to grow. Don't impose restrictions on the cloud. Just refactor what you need to for scale or enterprise grade or availability. >> And you would agree with that, is that true or is it problem you're solving? >> Well yeah, I mean it, what the cloud is doing is absolutely necessary. What the public cloud vendors are doing is absolutely necessary. But what's been missing is how to provide a consistent interface, especially to persistent data. And have it be available across different regions, and across different clouds. 'cause data is a highly localized thing in current architecture. It only exists as rendered by the storage system that you put it in. Whether that's a legacy thing like a NetApp or an Isilon or even a cloud data service. It's localized to a specific region of the cloud in which you put that. We have to delocalize data, and provide a consistent interface to it across all sites. That's high performance, local access, but to global data. >> And so Walmart earlier today described their, what we call Supercloud, they call it the Walmart cloud native platform. And they use this triplet model. They have AWS and Azure, no, oh sorry, no AWS. They have Azure and GCP and then on-prem, where all the VMs live. When you, you know, probe, it turns out that it's only stateless in the cloud. (John laughs) So, the state stuff- >> Well let's just admit it, there is no such thing as stateless, because even the application binaries and libraries are state. >> Well I'm happy that I'm hearing that. >> Yeah, okay. >> Because actually I have a lot of debate (indistinct). If you think about no software running on a (indistinct) machine is stateless. >> David: Exactly. >> This is something that was- >> David: And that's data that needs to be distributed and provided consistently >> (indistinct) >> Across all the clouds, >> And actually, it's a nonsense, but- >> Dave: So it's an illusion, okay. (group talks over each other) >> (indistinct) you guys talk about stateless. >> Well, see, people make the confusion between state and persistent state, okay. Persistent state it's a different thing. State is a different thing. So, but anyway, I want to go back to your point, because there's a lot of debate here. People are talking about data, some people are talking about logic, some people are talking about networking. In my opinion is this triplet, which is data logic and connectivity, that has equal importance. And actually depending on the application, can have the center of gravity moving towards data, moving towards what I call execution units or workloads. And connectivity is actually the most important part of it. >> David: (indistinct). >> Some people are saying move the logic towards the data, some other people, and you are saying actually, that no, you have to build a distributed data mesh. What I'm saying is actually, you have to consider all these three variables, all these vector in order to decide, based on application, what's the most important. Because sometimes- >> John: So the application chooses >> That's correct. >> Well it it's what operating systems were in the past, was principally the thing that runs and manages the jobs, the job scheduler, and the thing that provides your persistent data (indistinct). >> Okay. So we finally got operating system into the equation, thank you. (group laughs) >> Nelu: I actually have a PhD in operating system. >> Cause what we're talking about is an operating system. So forget platform or architecture, it's an operating environment. Let's use it as a general term. >> All right. I think that's about it for me. >> All right, let's take (indistinct). Nelu, I want ask you quick, 'cause I want to give a, 'cause I believe it's an operating system. I think it's going to be a reset, refactored. You wrote to me, "The model of Supercloud has to be open theoretical, has to satisfy the rigors of computer science, and customer requirements." So unique to today, if the OS is going to be refactored, it's not going to be, may or may not be Red Hat or somebody else. This new OS, obviously requirements are for customers too but is what's the computer science that is needed? Where are we, what's the missing? Where's the science in this shift? It's not your standard OS it's not like an- (group talks over each other) >> I would beg to differ. >> (indistinct) truly an operation environment. But the, if you think about, and make analogies, what you need when you design a distributed system, well you need an information model, yeah. You need to figure out how the data is located and distributed. You need a model for the execution units, and you need a way to describe the interactions between all these objects. And it is my opinion that we need to go deeper and formalize these operations in order to make a step forward. And when we design Supercloud, and design something that is better than the current HyperClouds. And actually that is when we design something better, you make a system more efficient and it's going to be better from the cost point of view, from the performance point of view. But we need to add some math into all this customer focus centering and I really admire AWS and their executive team focusing on the customer. But now it's time to go back and see, if we apply some computer science, if you try to formalize to build a theoretical model of cloud, can we build a system that is better than existing ones? >> So David, how do you- >> this is what I'm saying. >> That's a good question >> How do You see the operating system of a, or operating environment of a decentralized cloud? >> Well I think it's layered. I mean we have operating systems that can run systems quite efficiently. Linux has sort of one in the data center, but we're talking about a layer on top of that. And I think we're seeing the emergence of that. For example, on the job scheduling side of things, Kubernetes makes a really good example. You know, you break the workload into the most granular units of compute, the containerized microservice, and then you use a declarative model to state what is needed and give the system the degrees of freedom that it can choose how to instantiate it. Because the thing about these distributed systems, is that the complexity explodes, right? Running a piece of hardware, running a single server is not a problem, even with all the many cores and everything like that. It's when you start adding in the networking, and making it so that you have many of them. And then when it's going across whole different data centers, you know, so, at that level the way you solve this is not manually (group laughs) and not procedurally. You have to change the language so it's intent based, it's a declarative model, and what you're stating is what is intended, and you're leaving it to more advanced techniques, like machine learning to decide how to instantiate that service across the cluster, which is what Kubernetes does, or how to instantiate the data across the diverse storage infrastructure. And that's what we do. >> So that's a very good point because actually what has been neglected with HyperClouds is really optimization and automation. But in order to be able to do both of these things, you need, I'm going back and I'm stubborn, you need to have a mathematical model, a theoretical model because what does automation mean? It means that we have to put machines to do the work instead of us, and machines work with what? Formula, with algorithms, they don't work with services. So I think Supercloud is an opportunity to underscore the importance of optimization and automation- >> Totally agree. >> In HyperCloud, and actually by doing that, we can also have an interesting connotation. We are also contributing to save our planet, because if you think right now. we're consuming a lot of energy on this HyperClouds and also all this AI applications, and I think we can do better and build the same kind of application using less energy. >> So yeah, great point, love that call out, the- you know, Dave and I always joke about the old, 'cause we're old, we talk about, you know, (Nelu Laughs) old history, OS/2 versus DOS, okay, OS's, OS/2 is silly better, first threaded OS, DOS never went away. So how does legacy play into this conversation? Because I buy the theoretical, I love the conversation. Okay, I think it's an OS, totally see it that way myself. What's the blocker? Is there a legacy that drags it back? Is the anchor dragging from legacy? Is there a DOS OS/2 moment? Is there an opportunity to flip the script? This is- >> I think that's a perfect example of why we need to support the existing interfaces, Operating Systems, real operating systems like Linux, understands how to present data, it's called a file system, block devices, things that that plumb in there. And by, you know, going to a REST interface and S3 and telling people they have to rewrite their applications, you can't even consume your application binaries that way, the OS doesn't know how to pull that sort of thing. So we, to get to cloud, to get to the ability to host massive numbers of tenants within a centralized infrastructure, you know, we abandoned these lower level interfaces to the OS and we have to go back to that. It's the reason why DOS ultimately won, is it had the momentum of the install base. We're seeing the same thing here. Whatever it is, it has to be a real file system and not a come down file system >> Nelu, what's your reaction, 'cause you're in the theoretical bandwagon. Let's get your reaction. >> No, I think it's a good, I'll give, you made a good analogy between OS/2 and DOS, but I'll go even farther saying, if you think about the evolution operating system didn't stop the evolution of underlying microprocessors, hardware, and so on and so forth. On the contrary, it was a catalyst for that. So because everybody could develop their own hardware, without worrying that the applications on top of operating system are going to modify. The same thing is going to happen with Supercloud. You're going to have the AWSs, you're going to have the Azure and the the GCP continue to evolve in their own way proprietary. But if we create on top of it the right interface >> The open, this is why open is important. >> That's correct, because actually you're going to see sometime ago, everybody was saying, remember venture capitals were saying, "AWS killed the world, nobody's going to come." Now you see what Oracle is doing, and then you're going to see other players. >> It's funny, Amazon's trying to be more like Microsoft. Microsoft's trying to be more like Amazon and Google- Oracle's just trying to say they have cloud. >> That's, that's correct, (group laughs) so, my point is, you're going to see a multiplication of this HyperClouds and cloud technology. So, the system has to be open in order to accommodate what it is and what is going to come. Okay, so it's open. >> So the the legacy- so legacy is an opportunity, not a blocker in your mind. And you see- >> That's correct, I think we should allow them to continue to to to be their own actually. But maybe you're going to find a way to connect with it. >> Amazon's the processor, and they're on the 80 80 80 right? >> That's correct. >> You're saying you love people trying to get put to work. >> That's a good analogy. >> But, performance levels you say good luck, right? >> Well yeah, we have to be able to take traditional applications, high performance applications, those that consume file system and persistent data. Those things have to be able to run anywhere. You need to be able to put, put them onto, you know, more elastic infrastructure. So, we have to actually get cloud to where it lives up to its billing. >> And that's what you're solving for, with Hammerspace, >> That's what we're solving for, making it possible- >> Give me the bumper sticker. >> Solving for how do you have massive quantities of unstructured file data? At the end of the day, all data ultimately is unstructured data. Have that persistent data available, across any data center, within any cloud, within any region on-prem, at the edge. And have not just the same APIs, but have the exact same data sets, and not sucked over a straw remote, but at extreme high performance, local access. So how do you have local access to globally shared distributed data? And that's what we're doing. We are orchestrating data globally across all different forms of storage infrastructure, so you have a consistent access at the highest performance levels, at the lowest level innate built into the OS, how to consume it as (indistinct) >> So are you going into the- all the clouds and natively building in there, or are you off cloud? >> So This is software that can run on cloud instances and provide high performance file within the cloud. It can take file data that's on-prem. Again, it's software, it can run in virtual or on physical servers. And it abstracts the data from the existing storage infrastructure, and makes the data visible and consumable and orchestratable across any of it. >> And what's the elevator pitch for Cloud of Cloud, give that too. >> Well, Cloud of Clouds creates a theoretical model of cloud, and it describes every single object in the cloud. Where is data, execution units, and connectivity, with one single class of very simple object. And I can, I can give you (indistinct) >> And the problem that solves is what? >> The problem that solves is, it creates this mathematical model that is necessary in order to do other interesting things, such as optimization, using sata engines, using automation, applying ML for instance. Or deep learning to automate all this clouds, if you think about in the industrial field, we know how to manage and automate huge plants. Why wouldn't it do the same thing in cloud? It's the same thing you- >> That's what you mean by theoretical model. >> Nelu: That's correct. >> Lay out the architecture, almost the bones of skeleton or something, or, and then- >> That's correct, and then on top of it you can actually build a platform, You can create your services, >> when you say math, you mean you put numbers to it, you kind of index it. >> You quantify this thing and you apply mathematical- It's really about, I can disclose this thing. It's really about describing the cloud as a knowledge graph for every single object in the graph for node, an edge is a vector. And then once you have this model, then you can apply the field theory, and linear algebra to do operation with these vectors. And it's, this creates a very interesting opportunity to let the math do this thing for us. >> Okay, so what happens with hyperscale, or it's like AWS in your model. >> So in, in my model actually, >> Are they happy with this, or they >> I'm very happy with that. >> Will they be happy with you? >> We create an interface to every single HyperCloud. We actually, we don't need to interface with the thousands of APIs, but you know, if we have the 80 20 rule, and we map these APIs into this graph, and then every single operation that is done in this graph is done from the beginning, in an optimized manner and also automation ready. >> That's going to be great. David, I want us to go back to you before we close real quick. You've had a lot of experience, multiple ventures on the front end. You talked to a lot of customers who've been innovating. Where are the classic (indistinct)? Cause you, you used to sell and invent product around the old school enterprises with storage, you know that that trajectory storage is still critical to store the data. Where's the classic enterprise grade mindset right now? Those customers that were buying, that are buying storage, they're in the cloud, they're lifting and shifting. They not yet put the throttle on DevOps. When they look at this Supercloud thing, Are they like a deer in the headlights, or are they like getting it? What's the, what's the classic enterprise look like? >> You're seeing people at different stages of adoption. Some folks are trying to get to the cloud, some folks are trying to repatriate from the cloud, because they've realized it's better to own than to rent when you use a lot of it. And so people are at very different stages of the journey. But the one thing that's constant is that there's always change. And the change here has to do with being able to change the location where you're doing your computing. So being able to support traditional workloads in the cloud, being able to run things at the edge, and being able to rationalize where the data ought to exist, and with a declarative model, intent-based, business objective-based, be able to swipe a mouse and have the data get redistributed and positioned across different vendors, across different clouds, that, we're seeing that as really top of mind right now, because everybody's at some point on this journey, trying to go somewhere, and it involves taking their data with them. (John laughs) >> Guys, great conversation. Thanks so much for coming on, for John, Dave. Stay tuned, we got a great analyst power panel coming right up. More from Palo Alto, Supercloud 2. Be right back. (bouncy music)

>> From every ISV to solve the problems. You want there to be tools in place that you can use, either open source tools or whatever it is that help you build it. And slowly over time, that building will become easier and easier. So my question to you was, where do you see you playing? Do you see yourself playing to ISVs as a set of tools, which will make their life a lot easier and provide that work? >> Absolutely. >> If they don't have, so they don't have to do it. Or you're providing this for the end users? Or both? >> So it's a progression. If you go to the ISVs first, you're doomed to starved before you have time for that other option. >> Yeah. >> Right? So it's a question of phase, the phasing of it. And also if you go directly to end users, you can demonstrate the power of it and get the attention of the ISVs. I believe that the ISVs, especially those with the biggest footprints and the most, you know, coveted estates, they have already made massive investments at trying to solve decentralization of their software stack. And I believe that they have used it as a hook to try to move to a software as a service model and rope people into leasing their infrastructure. So if you look at the clouds that have been propped up by Autodesk or by Adobe, or you name the company, they are building proprietary makeshift solutions for decentralizing or hybrid clouding. Or maybe they're not even doing that at all and all they're is saying hey, if you want to get location agnosticness, then what you should just, is just move into our cloud. >> Right. >> And then they try to solve on the background how to decentralize it between different regions so they can have decent offerings in each region. But those who are more advanced have already made larger investments and will be more averse to, you know, throwing that stuff away, all of their makeshift machinery away, and using a platform that gives them high performance parallel, low level file system access, while at the same time having metadata-driven, you know, policy-based, intent-based orchestration to manage the diffusion of data across a decentralized infrastructure. They are not going to be as open because they've made such an investment and they're going to look at how do they monetize it. So what we have found with like the movie studios who are using us already, many of the app they're using, many of those software offerings, the ISVs have their own cloud that offers that software for the cloud. But what we got when I asked about this, 'cause I was dealt specifically into this question because I'm very interested to know how we're going to make that leap from end user upstream into the ISVs where I believe we need to, and they said, look, we cannot use these software ISV-specific SAS clouds for two reasons. Number one is we lose control of the data. We're giving it to them. That's security and other issues. And here you're talking about we're doing work for Disney, we're doing work for Netflix, and they're not going to let us put our data on those software clouds, on those SAS clouds. Secondly, in any reasonable pipeline, the data is shared by many different applications. We need to be agnostic as to the application. 'Cause the inputs to one application, you know, the output for one application provides the input to the next, and it's not necessarily from the same vendor. So they need to have a data platform that lets them, you know, go from one software stack, and you know, to run it on another. Because they might do the rendering with this and yet, they do the editing with that, and you know, et cetera, et cetera. So I think the further you go up the stack in the structured data and dedicated applications for specific functions in specific verticals, the further up the stack you go, the harder it is to justify a SAS offering where you're basically telling the end users you need to park all your data with us and then you can run your application in our cloud and get this. That ultimately is a dead end path versus having the data be open and available to many applications across this supercloud layer. >> Okay, so-- >> Is that making any sense? >> Yes, so if I could just ask a clarifying question. So, if I had to take Snowflake as an example, I think they're doing exactly what you're saying is a dead end, put everything into our proprietary system and then we'll figure out how to distribute it. >> Yeah. >> And and I think if you're familiar with Zhamak Dehghaniis' data mesh concept. Are you? >> A little bit, yeah. >> But in her model, Snowflake, a Snowflake warehouse is just a node on the mesh and that mesh is-- >> That's right. >> Ultimately the supercloud and you're an enabler of that is what I'm hearing. >> That's right. What they're doing up at the structured level and what they're talking about at the structured level we're doing at the underlying, unstructured level, which by the way has implications for how you implement those distributed database things. In other words, implementing a Snowflake on top of Hammerspace would have made building stuff like in the first place easier. It would allow you to easily shift and run the database engine anywhere. You still have to solve how to shard and distribute at the transaction layer above, so I'm not saying we're a substitute for what you need to do at the app layer. By the way, there is another example of that and that's Microsoft Office, right? It's one thing to share that, to have a file share where you can share all the docs. It's something else to have Word and PowerPoint, Excel know how to allow people to be simultaneously editing the same doc. That's always going to happen in the app layer. But not all applications need that level of, you know, in-app decentralization. You know, many of them, many workflows are pipelined, especially the ones that are very data intensive where you're doing drug discovery or you're doing rendering, or you're doing machine learning training. These things are human in the loop with large stages of processing across tens of thousands of cores. And I think that kind of data processing pipeline is what we're focusing on first. Not so much the Microsoft Office or the Snowflake, you know, parking a relational database because that takes a lot of application layer stuff and that's what they're good at. >> Right. >> But I think... >> Go ahead, sorry. >> Later entrance in these markets will find Hammerspace as a way to accelerate their work so they can focus more narrowly on just the stuff that's app-specific, higher level sharing in the app. >> Yes, Snowflake founders-- >> I think it might be worth mentioning also, just keep this confidential guys, but one of our customers is Blue Origin. And one of the things that we have found is kind of the point of what you're talking about with our customers. They're needing to build this and since it's not commercially available or they don't know where to look for it to be commercially available, they're all building themselves. So this layer is needed. And Blue is just one of the examples of quite a few we're now talking to. And like manufacturing, HPC, research where they're out trying to solve this problem with their own scripting tools and things like that. And I just, I don't know if there's anything you want to add, David, but you know, but there's definitely a demand here and customers are trying to figure out how to solve it beyond what Hammerspace is doing. Like the need is so great that they're just putting developers on trying to do it themselves. >> Well, and you know, Snowflake founders, they didn't have a Hammerspace to lean on. But, one of the things that's interesting about supercloud is we feel as though industry clouds will emerge, that as part of company's digital transformations, they will, you know, every company's a software company, they'll begin to build their own clouds and they will be able to use a Hammerspace to do that. >> A super pass layer. >> Yes. It's really, I don't know if David's speaking, I don't want to speak over him, but we can't hear you. May be going through a bad... >> Well, a regional, regional talks that make that possible. And so they're doing these render farms and editing farms, and it's a cloud-specific to the types of workflows in the median entertainment world. Or clouds specifically to workflows in the chip design world or in the drug and bio and life sciences exploration world. There are large organizations that are kind of a blend of end users, like the Broad, which has their own kind of cloud where they're asking collaborators to come in and work with them. So it starts to even blur who's an end user versus an ISV. >> Yes. >> Right? When you start talking about the massive data is the main gravity is to having lots of people participate. >> Yep, and that's where the value is. And that's where the value is. And this is a megatrend that we see. And so it's really important for us to get to the point of what is and what is not a supercloud and, you know, that's where we're trying to evolve. >> Let's talk about this for a second 'cause I want to, I want to challenge you on something and it's something that I got challenged on and it has led me to thinking differently than I did at first, which Molly can attest to. Okay? So, we have been looking for a way to talk about the concept of cloud of utility computing, run anything anywhere that isn't addressed in today's realization of cloud. 'Cause today's cloud is not run anything anywhere, it's quite the opposite. You park your data in AWS and that's where you run stuff. And you pretty much have to. Same with with Azure. They're using data gravity to keep you captive there, just like the old infrastructure guys did. But now it's even worse because it's coupled back with the software to some degree, as well. And you have to use their storage, networking, and compute. It's not, I mean it fell back to the mainframe era. Anyhow, so I love the concept of supercloud. By the way, I was going to suggest that a better term might be hyper cloud since hyper speaks to the multidimensionality of it and the ability to be in a, you know, be in a different dimension, a different plane of existence kind of thing like hyperspace. But super and hyper are somewhat synonyms. I mean, you have hyper cars and you have super cars and blah, blah, blah. I happen to like hyper maybe also because it ties into the whole Hammerspace notion of a hyper-dimensional, you know, reality, having your data centers connected by a wormhole that is Hammerspace. But regardless, what I got challenged on is calling it something different at all versus simply saying, this is what cloud has always meant to be. This is the true cloud, this is real cloud, this is cloud. And I think back to what happened, you'll remember, at Fusion IO we talked about IO memory and we did that because people had a conceptualization of what an SSD was. And an SSD back then was low capacity, low endurance, made to go military, aerospace where things needed to be rugged but was completely useless in the data center. And we needed people to imagine this thing as being able to displace entire SAND, with the kind of capacity density, performance density, endurance. And so we talked IO memory, we could have said enterprise SSD, and that's what the industry now refers to for that concept. What will people be saying five and 10 years from now? Will they simply say, well this is cloud as it was always meant to be where you are truly able to run anything anywhere and have not only the same APIs, but you're same data available with high performance access, all forms of access, block file and object everywhere. So yeah. And I wonder, and this is just me throwing it out there, I wonder if, well, there's trade offs, right? Giving it a new moniker, supercloud, versus simply talking about how cloud is always intended to be and what it was meant to be, you know, the real cloud or true cloud, there are trade-offs. By putting a name on it and branding it, that lets people talk about it and understand they're talking about something different. But it also is that an affront to people who thought that that's what they already had. >> What's different, what's new? Yes, and so we've given a lot of thought to this. >> Right, it's like you. >> And it's because we've been asked that why does the industry need a new term, and we've tried to address some of that. But some of the inside baseball that we haven't shared is, you remember the Web 2.0, back then? >> Yep. >> Web 2.0 was the same thing. And I remember Tim Burners Lee saying, "Why do we need Web 2.0? "This is what the Web was always supposed to be." But the truth is-- >> I know, that was another perfect-- >> But the truth is it wasn't, number one. Number two, everybody hated the Web 2.0 term. John Furrier was actually in the middle of it all. And then it created this groundswell. So one of the things we wrote about is that supercloud is an evocative term that catalyzes debate and conversation, which is what we like, of course. And maybe that's self-serving. But yeah, HyperCloud, Metacloud, super, meaning, it's funny because super came from Latin supra, above, it was never the superlative. But the superlative was a convenient byproduct that caused a lot of friction and flack, which again, in the media business is like a perfect storm brewing. >> The bad thing to have to, and I think you do need to shake people out of their, the complacency of the limitations that they're used to. And I'll tell you what, the fact that you even have the terms hybrid cloud, multi-cloud, private cloud, edge computing, those are all just referring to the different boundaries that isolate the silo that is the current limited cloud. >> Right. >> So if I heard correctly, what just, in terms of us defining what is and what isn't in supercloud, you would say traditional applications which have to run in a certain place, in a certain cloud can't run anywhere else, would be the stuff that you would not put in as being addressed by supercloud. And over time, you would want to be able to run the data where you want to and in any of those concepts. >> Or even modern apps, right? Or even modern apps that are siloed in SAS within an individual cloud, right? >> So yeah, I guess it's twofold. Number one, if you're going at the high application layers, there's lots of ways that you can give the appearance of anything running anywhere. The ISV, the SAS vendor can engineer stuff to have the ability to serve with low enough latency to different geographies, right? So if you go too high up the stack, it kind of loses its meaning because there's lots of different ways to make due and give the appearance of omni-presence of the service. Okay? As you come down more towards the platform layer, it gets harder and harder to mask the fact that supercloud is something entirely different than just a good regionally-distributed SAS service. So I don't think you, I don't think you can distinguish supercloud if you go too high up the stack because it's just SAS, it's just a good SAS service where the SAS vendor has done the hard work to give you low latency access from different geographic regions. >> Yeah, so this is one of the hardest things, David. >> Common among them. >> Yeah, this is really an important point. This is one of the things I've had the most trouble with is why is this not just SAS? >> So you dilute your message when you go up to the SAS layer. If you were to focus most of this around the super pass layer, the how can you host applications and run them anywhere and not host this, not run a service, not have a service available everywhere. So how can you take any application, even applications that are written, you know, in a traditional legacy data center fashion and be able to run them anywhere and have them have their binaries and their datasets and the runtime environment and the infrastructure to start them and stop them? You know, the jobs, the, what the Kubernetes, the job scheduler? What we're really talking about here, what I think we're really talking about here is building the operating system for a decentralized cloud. What is the operating system, the operating environment for a decentralized cloud? Where you can, and that the main two functions of an operating system or an operating environment are the process scheduler, the thing that's scheduling what is running where and when and so forth, and the file system, right? The thing that's supplying a common view and access to data. So when we talk about this, I think that the strongest argument for supercloud is made when you go down to the platform layer and talk of it, talk about it as an operating environment on which you can run all forms of applications. >> Would you exclude--? >> Not a specific application that's been engineered as a SAS. (audio distortion) >> He'll come back. >> Are you there? >> Yeah, yeah, you just cut out for a minute. >> I lost your last statement when you broke up. >> We heard you, you said that not the specific application. So would you exclude Snowflake from supercloud? >> Frankly, I would. I would. Because, well, and this is kind of hard to do because Snowflake doesn't like to, Frank doesn't like to talk about Snowflake as a SAS service. It has a negative connotation. >> But it is. >> I know, we all know it is. We all know it is and because it is, yes, I would exclude them. >> I think I actually have him on camera. >> There's nothing in common. >> I think I have him on camera or maybe Benoit as saying, "Well, we are a SAS." I think it's Slootman. I think I said to Slootman, "I know you don't like to say you're a SAS." And I think he said, "Well, we are a SAS." >> Because again, if you go to the top of the application stack, there's any number of ways you can give it location agnostic function or you know, regional, local stuff. It's like let's solve the location problem by having me be your one location. How can it be decentralized if you're centralizing on (audio distortion)? >> Well, it's more decentralized than if it's all in one cloud. So let me actually, so the spectrum. So again, in the spirit of what is and what isn't, I think it's safe to say Hammerspace is supercloud. I think there's no debate there, right? Certainly among this crowd. And I think we can all agree that Dell, Dell Storage is not supercloud. Where it gets fuzzy is this Snowflake example or even, how about a, how about a Cohesity that instantiates its stack in different cloud regions in different clouds, and synchronizes, however magic sauce it does that. Is that a supercloud? I mean, so I'm cautious about having too strict of a definition 'cause then only-- >> Fair enough, fair enough. >> But I could use your help and thoughts on that. >> So I think we're talking about two different spectrums here. One is the spectrum of platform to application-specific. As you go up the application stack and it becomes this specific thing. Or you go up to the more and more structured where it's serving a specific application function where it's more of a SAS thing. I think it's harder to call a SAS service a supercloud. And I would argue that the reason there, and what you're lacking in the definition is to talk about it as general purpose. Okay? Now, that said, a data warehouse is general purpose at the structured data level. So you could make the argument for why Snowflake is a supercloud by saying that it is a general purpose platform for doing lots of different things. It's just one at a higher level up at the structured data level. So one spectrum is the high level going from platform to, you know, unstructured data to structured data to very application-specific, right? Like a specific, you know, CAD/CAM mechanical design cloud, like an Autodesk would want to give you their cloud for running, you know, and sharing CAD/CAM designs, doing your CAD/CAM anywhere stuff. Well, the other spectrum is how well does the purported supercloud technology actually live up to allowing you to run anything anywhere with not just the same APIs but with the local presence of data with the exact same runtime environment everywhere, and to be able to correctly manage how to get that runtime environment anywhere. So a Cohesity has some means of running things in different places and some means of coordinating what's where and of serving diff, you know, things in different places. I would argue that it is a very poor approximation of what Hammerspace does in providing the exact same file system with local high performance access everywhere with metadata ability to control where the data is actually instantiated so that you don't have to wait for it to get orchestrated. But even then when you do have to wait for it, it happens automatically and so it's still only a matter of, well, how quick is it? And on the other end of the spectrum is you could look at NetApp with Flexcache and say, "Is that supercloud?" And I would argue, well kind of because it allows you to run things in different places because it's a cache. But you know, it really isn't because it presumes some central silo from which you're cacheing stuff. So, you know, is it or isn't it? Well, it's on a spectrum of exactly how fully is it decoupling a runtime environment from specific locality? And I think a cache doesn't, it stretches a specific silo and makes it have some semblance of similar access in other places. But there's still a very big difference to the central silo, right? You can't turn off that central silo, for example. >> So it comes down to how specific you make the definition. And this is where it gets kind of really interesting. It's like cloud. Does IBM have a cloud? >> Exactly. >> I would say yes. Does it have the kind of quality that you would expect from a hyper-scale cloud? No. Or see if you could say the same thing about-- >> But that's a problem with choosing a name. That's the problem with choosing a name supercloud versus talking about the concept of cloud and how true up you are to that concept. >> For sure. >> Right? Because without getting a name, you don't have to draw, yeah. >> I'd like to explore one particular or bring them together. You made a very interesting observation that from a enterprise point of view, they want to safeguard their store, their data, and they want to make sure that they can have that data running in their own workflows, as well as, as other service providers providing services to them for that data. So, and in in particular, if you go back to, you go back to Snowflake. If Snowflake could provide the ability for you to have your data where you wanted, you were in charge of that, would that make Snowflake a supercloud? >> I'll tell you, in my mind, they would be closer to my conceptualization of supercloud if you can instantiate Snowflake as software on your own infrastructure, and pump your own data to Snowflake that's instantiated on your own infrastructure. The fact that it has to be on their infrastructure or that it's on their, that it's on their account in the cloud, that you're giving them the data and they're, that fundamentally goes against it to me. If they, you know, they would be a pure, a pure plate if they were a software defined thing where you could instantiate Snowflake machinery on the infrastructure of your choice and then put your data into that machinery and get all the benefits of Snowflake. >> So did you see--? >> In other words, if they were not a SAS service, but offered all of the similar benefits of being, you know, if it were a service that you could run on your own infrastructure. >> So did you see what they announced, that--? >> I hope that's making sense. >> It does, did you see what they announced at Dell? They basically announced the ability to take non-native Snowflake data, read it in from an object store on-prem, like a Dell object store. They do the same thing with Pure, read it in, running it in the cloud, and then push it back out. And I was saying to Dell, look, that's fine. Okay, that's interesting. You're taking a materialized view or an extended table, whatever you're doing, wouldn't it be more interesting if you could actually run the query locally with your compute? That would be an extension that would actually get my attention and extend that. >> That is what I'm talking about. That's what I'm talking about. And that's why I'm saying I think Hammerspace is more progressive on that front because with our technology, anybody who can instantiate a service, can make a service. And so I, so MSPs can use Hammerspace as a way to build a super pass layer and host their clients on their infrastructure in a cloud-like fashion. And their clients can have their own private data centers and the MSP or the public clouds, and Hammerspace can be instantiated, get this, by different parties in these different pieces of infrastructure and yet linked together to make a common file system across all of it. >> But this is data mesh. If I were HPE and Dell it's exactly what I'd be doing. I'd be working with Hammerspace to create my own data. I'd work with Databricks, Snowflake, and any other-- >> Data mesh is a good way to put it. Data mesh is a good way to put it. And this is at the lowest level of, you know, the underlying file system that's mountable by the operating system, consumed as a real file system. You can't get lower level than that. That's why this is the foundation for all of the other apps and structured data systems because you need to have a data mesh that can at least mesh the binary blob. >> Okay. >> That hold the binaries and that hold the datasets that those applications are running. >> So David, in the third week of January, we're doing supercloud 2 and I'm trying to convince John Furrier to make it a data slash data mesh edition. I'm slowly getting him to the knothole. I would very much, I mean you're in the Bay Area, I'd very much like you to be one of the headlines. As Zhamak Dehghaniis going to speak, she's the creator of Data Mesh, >> Sure. >> I'd love to have you come into our studio as well, for the live session. If you can't make it, we can pre-record. But you're right there, so I'll get you the dates. >> We'd love to, yeah. No, you can count on it. No, definitely. And you know, we don't typically talk about what we do as Data Mesh. We've been, you know, using global data environment. But, you know, under the covers, that's what the thing is. And so yeah, I think we can frame the discussion like that to line up with other, you know, with the other discussions. >> Yeah, and Data Mesh, of course, is one of those evocative names, but she has come up with some very well defined principles around decentralized data, data as products, self-serve infrastructure, automated governance, and and so forth, which I think your vision plugs right into. And she's brilliant. You'll love meeting her. >> Well, you know, and I think.. Oh, go ahead. Go ahead, Peter. >> Just like to work one other interface which I think is important. How do you see yourself and the open source? You talked about having an operating system. Obviously, Linux is the operating system at one level. How are you imagining that you would interface with cost community as part of this development? >> Well, it's funny you ask 'cause my CTO is the kernel maintainer of the storage networking stack. So how the Linux operating system perceives and consumes networked data at the file system level, the network file system stack is his purview. He owns that, he wrote most of it over the last decade that he's been the maintainer, but he's the gatekeeper of what goes in. And we have leveraged his abilities to enhance Linux to be able to use this decentralized data, in particular with decoupling the control plane driven by metadata from the data access path and the many storage systems on which the data gets accessed. So this factoring, this splitting of control plane from data path, metadata from data, was absolutely necessary to create a data mesh like we're talking about. And to be able to build this supercloud concept. And the highways on which the data runs and the client which knows how to talk to it is all open source. And we have, we've driven the NFS 4.2 spec. The newest NFS spec came from my team. And it was specifically the enhancements needed to be able to build a spanning file system, a data mesh at a file system level. Now that said, our file system itself and our server, our file server, our data orchestration, our data management stuff, that's all closed source, proprietary Hammerspace tech. But the highways on which the mesh connects are actually all open source and the client that knows how to consume it. So we would, honestly, I would welcome competitors using those same highways. They would be at a major disadvantage because we kind of built them, but it would still be very validating and I think only increase the potential adoption rate by more than whatever they might take of the market. So it'd actually be good to split the market with somebody else to come in and share those now super highways for how to mesh data at the file system level, you know, in here. So yeah, hopefully that answered your question. Does that answer the question about how we embrace the open source? >> Right, and there was one other, just that my last one is how do you enable something to run in every environment? And if we take the edge, for example, as being, as an environment which is much very, very compute heavy, but having a lot less capability, how do you do a hold? >> Perfect question. Perfect question. What we do today is a software appliance. We are using a Linux RHEL 8, RHEL 8 equivalent or a CentOS 8, or it's, you know, they're all roughly equivalent. But we have bundled and a software appliance which can be instantiated on bare metal hardware on any type of VM system from VMware to all of the different hypervisors in the Linux world, to even Nutanix and such. So it can run in any virtualized environment and it can run on any cloud instance, server instance in the cloud. And we have it packaged and deployable from the marketplaces within the different clouds. So you can literally spin it up at the click of an API in the cloud on instances in the cloud. So with all of these together, you can basically instantiate a Hammerspace set of machinery that can offer up this file system mesh. like we've been using the terminology we've been using now, anywhere. So it's like being able to take and spin up Snowflake and then just be able to install and run some VMs anywhere you want and boom, now you have a Snowflake service. And by the way, it is so complete that some of our customers, I would argue many aren't even using public clouds at all, they're using this just to run their own data centers in a cloud-like fashion, you know, where they have a data service that can span it all. >> Yeah and to Molly's first point, we would consider that, you know, cloud. Let me put you on the spot. If you had to describe conceptually without a chalkboard what an architectural diagram would look like for supercloud, what would you say? >> I would say it's to have the same runtime environment within every data center and defining that runtime environment as what it takes to schedule the execution of applications, so job scheduling, runtime stuff, and here we're talking Kubernetes, Slurm, other things that do job scheduling. We're talking about having a common way to, you know, instantiate compute resources. So a global compute environment, having a common compute environment where you can instantiate things that need computing. Okay? So that's the first part. And then the second is the data platform where you can have file block and object volumes, and have them available with the same APIs in each of these distributed data centers and have the exact same data omnipresent with the ability to control where the data is from one moment to the next, local, where all the data is instantiate. So my definition would be a common runtime environment that's bifurcate-- >> Oh. (attendees chuckling) We just lost them at the money slide. >> That's part of the magic makes people listen. We keep someone on pin and needles waiting. (attendees chuckling) >> That's good. >> Are you back, David? >> I'm on the edge of my seat. Common runtime environment. It was like... >> And just wait, there's more. >> But see, I'm maybe hyper-focused on the lower level of what it takes to host and run applications. And that's the stuff to schedule what resources they need to run and to get them going and to get them connected through to their persistence, you know, and their data. And to have that data available in all forms and have it be the same data everywhere. On top of that, you could then instantiate applications of different types, including relational databases, and data warehouses and such. And then you could say, now I've got, you know, now I've got these more application-level or structured data-level things. I tend to focus less on that structured data level and the application level and am more focused on what it takes to host any of them generically on that super pass layer. And I'll admit, I'm maybe hyper-focused on the pass layer and I think it's valid to include, you know, higher levels up the stack like the structured data level. But as soon as you go all the way up to like, you know, a very specific SAS service, I don't know that you would call that supercloud. >> Well, and that's the question, is there value? And Marianna Tessel from Intuit said, you know, we looked at it, we did it, and it just, it was actually negative value for us because connecting to all these separate clouds was a real pain in the neck. Didn't bring us any additional-- >> Well that's 'cause they don't have this pass layer underneath it so they can't even shop around, which actually makes it hard to stand up your own SAS service. And ultimately they end up having to build their own infrastructure. Like, you know, I think there's been examples like Netflix moving away from the cloud to their own infrastructure. Basically, if you're going to rent it for more than a few months, it makes sense to build it yourself, if it's at any kind of scale. >> Yeah, for certain components of that cloud. But if the Goldman Sachs came to you, David, and said, "Hey, we want to collaborate and we want to build "out a cloud and essentially build our SAS system "and we want to do that with Hammerspace, "and we want to tap the physical infrastructure "of not only our data centers but all the clouds," then that essentially would be a SAS, would it not? And wouldn't that be a Super SAS or a supercloud? >> Well, you know, what they may be using to build their service is a supercloud, but their service at the end of the day is just a SAS service with global reach. Right? >> Yeah. >> You know, look at, oh shoot. What's the name of the company that does? It has a cloud for doing bookkeeping and accounting. I forget their name, net something. NetSuite. >> NetSuite. NetSuite, yeah, Oracle. >> Yeah. >> Yep. >> Oracle acquired them, right? Is NetSuite a supercloud or is it just a SAS service? You know? I think under the covers you might ask are they using supercloud under the covers so that they can run their SAS service anywhere and be able to shop the venue, get elasticity, get all the benefits of cloud in the, to the benefit of their service that they're offering? But you know, folks who consume the service, they don't care because to them they're just connecting to some endpoint somewhere and they don't have to care. So the further up the stack you go, the more location-agnostic it is inherently anyway. >> And I think it's, paths is really the critical layer. We thought about IAS Plus and we thought about SAS Minus, you know, Heroku and hence, that's why we kind of got caught up and included it. But SAS, I admit, is the hardest one to crack. And so maybe we exclude that as a deployment model. >> That's right, and maybe coming down a level to saying but you can have a structured data supercloud, so you could still include, say, Snowflake. Because what Snowflake is doing is more general purpose. So it's about how general purpose it is. Is it hosting lots of other applications or is it the end application? Right? >> Yeah. >> So I would argue general purpose nature forces you to go further towards platform down-stack. And you really need that general purpose or else there is no real distinguishing. So if you want defensible turf to say supercloud is something different, I think it's important to not try to wrap your arms around SAS in the general sense. >> Yeah, and we've kind of not really gone, leaned hard into SAS, we've just included it as a deployment model, which, given the constraints that you just described for structured data would apply if it's general purpose. So David, super helpful. >> Had it sign. Define the SAS as including the hybrid model hold SAS. >> Yep. >> Okay, so with your permission, I'm going to add you to the list of contributors to the definition. I'm going to add-- >> Absolutely. >> I'm going to add this in. I'll share with Molly. >> Absolutely. >> We'll get on the calendar for the date. >> If Molly can share some specific language that we've been putting in that kind of goes to stuff we've been talking about, so. >> Oh, great. >> I think we can, we can share some written kind of concrete recommendations around this stuff, around the general purpose, nature, the common data thing and yeah. >> Okay. >> Really look forward to it and would be glad to be part of this thing. You said it's in February? >> It's in January, I'll let Molly know. >> Oh, January. >> What the date is. >> Excellent. >> Yeah, third week of January. Third week of January on a Tuesday, whatever that is. So yeah, we would welcome you in. But like I said, if it doesn't work for your schedule, we can prerecord something. But it would be awesome to have you in studio. >> I'm sure with this much notice we'll be able to get something. Let's make sure we have the dates communicated to Molly and she'll get my admin to set it up outside so that we have it. >> I'll get those today to you, Molly. Thank you. >> By the way, I am so, so pleased with being able to work with you guys on this. I think the industry needs it very bad. They need something to break them out of the box of their own mental constraints of what the cloud is versus what it's supposed to be. And obviously, the more we get people to question their reality and what is real, what are we really capable of today that then the more business that we're going to get. So we're excited to lend the hand behind this notion of supercloud and a super pass layer in whatever way we can. >> Awesome. >> Can I ask you whether your platforms include ARM as well as X86? >> So we have not done an ARM port yet. It has been entertained and won't be much of a stretch. >> Yeah, it's just a matter of time. >> Actually, entertained doing it on behalf of NVIDIA, but it will absolutely happen because ARM in the data center I think is a foregone conclusion. Well, it's already there in some cases, but not quite at volume. So definitely will be the case. And I'll tell you where this gets really interesting, discussion for another time, is back to my old friend, the SSD, and having SSDs that have enough brains on them to be part of that fabric. Directly. >> Interesting. Interesting. >> Very interesting. >> Directly attached to ethernet and able to create a data mesh global file system, that's going to be really fascinating. Got to run now. >> All right, hey, thanks you guys. Thanks David, thanks Molly. Great to catch up. Bye-bye. >> Bye >> Talk to you soon.