>> Announcer: From theCUBE studios (upbeat music) in Palo Alto in Boston, connecting with thought leaders all around the world, this is a CUBE Conversation. >> Hi, and welcome to a special CUBE conversation. I'm Stu Miniman, coming to you from our Boston area studio, and really happy to welcome to the program to dig into some of the latest on what's going on in the multi-cloud ecosystem. First of all, coming back to the program, not too far from where I'm sitting, Joe CaraDonna. He is the Vice President of Engineering Technologies, with Dell Technologies, and joining him, someone he knows quite well, is Rich Sanzi, who's Vice President of Engineering at Google Cloud. Gentlemen, thanks so much for joining. >> Great to be here, Stu. >> Thank you. >> All right, so Joe, we've been watching Dell Technologies, how the cloud portfolio and solution has been maturing, and working with the ecosystem. Maybe set the table for us, what's Dell doing with cloud? Why are we sitting here with the ? >> Well, we're here to talk about our OneFS for Google Cloud offering. We did something really special with Google here. We brought together the power and scale of our OneFS file system, along with the economics and the simplicity of public cloud, and together, I think, what we did is define a new standard for scalable file in public cloud, where we have a game-changing performance and capacity. We have a full range of enterprise-grade data management capabilities, and we enable real hybrid cloud, and open up new use cases for our customers. >> Excellent, thanks Joe for setting the table on that. Rich, let's pull you into the conversation. Before we go into the Google thing, give us a little bit about your background. You've been in storage, as I hinted at. You worked with Joe before, and tell us about your role inside of Google. >> Yeah, so I actually joined Google a few years ago, responsible for storage, and storage for all of Google, in addition to Google Cloud. And then, you know, big company things. We've been growing rapidly, and an opportunity opened up where I could be much more engaged on the Compute side, and so I'm responsible for Compute, the IaaS infrastructure for Google Cloud Engine. So it's my pleasure to be here and support Joe and Dell Technologies in the launch of OneFS on Google Cloud. >> Yeah, Rich, I'd like to come back to you on something, 'cause when you look at cloud, for many years it was cloud versus, you know, taking over the world, destroying everything before it. And especially, you look at Compute, or storage specifically, people have a little bit of a hard time wrapping their heads around, where my application lives. Does it just live one place? Are my applications going a little bit hybrid there? I look back, you know the disclosure, I worked at EMP for years. You know that storage is complicated and diverse, that's why we have file, block, and object. We have lots of different types of solutions out there. There's never been a silver bullet that says, "Okay, 90% of the people can use this one thing "for everything." So Rich, let's start with you. Cloud definitely has changed the discussion of storage, but I feel like I've seen the enterprise solutions looking more like the hyperscalers, and the hyperscale solution blurring the lines with what was traditionally happening in the data center. Do you agree with some of that? >> Oh yeah, absolutely. I think it's really nice when you control the horizontal and the vertical, and you can adapt your application stack, but that's just not the reality where we are today. The reality is that a cloud vendor, working with customers who bring their workloads in the cloud, have to be able to support all of the best-in-class types of storage that people are using. You're absolutely right, we're using cloud, or sorry, we're using objects, we're using block, we're using file. One of the great pieces of this, is that in the file space, you really need scalable file to go along with your scalable compute. >> Excellent, so-- >> Yeah, and I'll just add-- >> Please, go ahead Joe. >> Yeah, I mean, our customers, for a long time, have been asking, our Isilon customers in particular, asking for a long time to bring this type of capability to the cloud. They want the scalability of the elastic compute in the GPUs. They also want the OpEx model, right? And they want to be able to bring the high performance compute workloads to the cloud, but they need a scalable file system that can keep up with the demand, and that's what we set out to solve for. >> Excellent, so Joe you mentioned that the Isilon piece. You know, we've watched what has happened with that. You know, Isilon has always been software at the core and highly scalable, so we'd like you both, Joe you teed it up there, but Rich, why is this important for Google Cloud customers, and how's it different from, maybe, how they were doing things in the past? >> Well, I think one of the things that I think I'm really excited about, is that this enables customers to leverage the cloud, and not make a ton of changes on their server side. So it really allows them to preserve their investment, and their applications, and the way that they think about storage, and the way they think about how that scales and performs. So that, for me, is a, let's make it easy for customers to consume cloud, rather than make it a hurdle, and that's my view. >> Yeah, and Joe, help frame this for us a bit. You know, we watched Dell Technologies recently had the Power Store announcement. A lot of discussion about cloud native architectures, moving to micro-services. Google's one of the earliest and most prominent examples of innerized architectures out there. So, where does the file solution fit in this whole discussion that customers have about modernization of their applications, and the journey that they're going on? >> Yeah, well, not all applications lend themselves well to object. They need file semantics, as well as the performance characteristics that come along with that, in terms of throughput and latencies. But even beyond that, what our customer's looking for is the data management capability, right? Whether it's snapshots, or the multi-protocol data access for NFS, or SMB, or even HDFS. And they're looking for replication, native replication, so they can have their Isilon systems in the data center, replicate their data directly into the file service of the cloud so they can actually operate on that data, and then there's things that we take for granted now, at least in the data center, of that high availability and that high durability, that storage arrays deliver. So, it's a combination of things that make it attractive for customers, that open up these new workloads, especially in terms of a high performance compute. >> Excellent, you talked a bit about some of the reasons why customers wouldn't want file. Of course, scale is one of those things we've been talking about for many years. Scale means different things to many people. There's few companies that know scale better than Google, so Rich, talk a little bit about scalability, performance, what these types of evolutions mean, and what you're hearing from customers. >> Certainly from a scale perspective, things like objects and object store is super scalable. It's also, you know, requires application changes, to really make use of. Customers are really looking for scalable solutions that enable them to bring their existing applications to cloud, and not have to make a ton of changes to it. That's one of the things I think is great about the Dell offering, is that it is a full-fidelity solution that has the performance and scale of what customers are expecting from their on-premise, and then when we wire that up with the Google network into our Google Cloud compute regions, we get very high performance, and very high fidelity, low latency as a result. We think that that removes potential headaches that customers may have when they bring big applications in the HBC space, and related high performance computing space in the cloud. >> Great, and Joe, is all this available now? Tell us a little bit about availability. What do you expect the demand to be for this solution? >> Well, I expect the demand to be great, right? The kind of workloads we're talking about here cut across a wide range of verticals. So everything from whether it's like sciences for genomics research, oil and gas for seismic data processing, media and entertainment for video editing and rendering, or even finishing, automotive telemetry data that requires processing and scale, and EDA. So, I think it hits upon a wide variety of use cases and verticals, and we've even structured our pricing and our tiers to make it more accessible for use cases from high performance, all the way down to even archival. >> So, maybe just to clarify, this is GA today? >> Yeah, yes, it is GA. (laughs) >> Okay, excellent. >> Beta is behind 'em. >> Appreciate that, and how does, you mentioned flexibility on pricing. How much of this is what's available from Google, what's available from Dell? How does that relationship and go-to-market work together? >> Yeah, well it's a native service in Google. You can provision directly from the Google Portal. You can manage your file systems directly from the Google Portal, and the billing is integrated. So you get one bill from Google, whether it's for our OneFS file service, or any of Google's native services. >> Excellent, Rich, we'd love to hear, talk about from the Google side, the ecosystem. I know last year, I was at the Google Next event, really saw strong demand from the partner community. They're looking to work with Google, many have worked with Google for many years. What kind of feedback have you been getting and how this fits into the overall solution? >> So, from a partner perspective, one of the things that we really want to enable our partners, is to bring their services onto our platform, and to integrate them tightly as if they were a Google offering, and that's so things like the integrated billing, the provisioning from the Google Portal, things like that are core tenets for us for helping our customers and our partners' customers easily consume services in the cloud. So, sort of one of the P-zero requirements, from my perspective, for our product offering here, was that in fact it was just integrated into the Google Cloud platform, and that it would be discoverable and easily usable by customers. So I think that enables partners to deliver a first-class service on our platform. >> Yeah, I mean, Rich, absolutely. Some of the feedback I've gotten from the ecosystem, is, how do they put it? They say, "Google kind of puts you through the ringer. "By the time you get through that, "it is going to work." And of course, we know, Google's doing that to make sure that there are good, reliable, strong services by the time the end customer gets them. All right, Joe-- >> Yes, and-- >> (laughs) Go ahead, yeah. >> I was going to say, you know, delivering these services, and delivering them reliably, it's a multi-company partnership, but we understand that at the end of the day, the customer wants to be assured that there is, they have one contact for problems with the service, and so that's where Google very much wants to be that primary contact, 'cause who knows where the issues could be. Are they in the data center, or are they in the network, or are they on the customer side? We feel responsibility to front (audio distorts). >> Yeah, absolutely. So, Joe, I guess, final thing for you. Talk about the Dell Technologies Google Cloud relationship, why that's important, what differentiates it from some of the many other partnerships that Dell has. >> Yeah, sure, before I touch on that, I want to talk about, you mentioned scale, and scale means different things to different people. And when we're talking about scale here, capacity's one element of that, and we certainly scale that way, but performance is the other way. And ESG did a performance study on the OneFS file service that we're offering, and they fired up the biozone benchmark, which fired up over 1000 cores in Google, running NFS load to the file system. They sized the file system at 2 petabytes, which seems large, and it is, but you can scale much larger than that with our service, and their results on throughput was 200 gigabytes per second on the read, and 100 gigabytes per second on the write. Now, these are game changing numbers, right? It's numbers like that that enable compute-intensive, high performance workloads in Google Cloud, and we're opening that up. And it's also important to note that this is a scalable file system, so if you want to double those throughput numbers, you just double the capacity of your file system. So that's the power of scale that we're delivering here. And our file system can scale up to 50 petabytes, so a lot of runway there. As far as the partnership with Google goes, I mean, Google's been great. Their infrastructure is amazing. In order to hit those kind of performance numbers, your head goes to compute and the file system, but there's also a network in there, and to hit those kind of numbers, Google had to supply a two terabyte per second network, and they were able to supply the compute and the network with ease, and without hiccup. So it's together that we're solving for the compute, network, and storage equation, and that we can deliver a holistic solution. And lastly, I would just point out, the engineering teams work great bringing that cloud native experience into that Google Portal, really simplifying user experience. So, they can provision and manage the systems directly from the Portal, as well as unifying the billing. So I think the partnership's been great, and it's going to be interesting to see how our customers use the service to accelerate their cloud journey. >> Well, Joe and Rich, thank you so much for the updates. Congratulations on GA of this, and definitely look forward to hearing the customer journeys as they go on. >> Thank you, Stu. >> All right, thank you. And Rich, thank you for your partnership. >> Yeah, your welcome, Joe. Thank you, as well. >> All right, be sure to check out thecube.net for all the coverage, the virtual events that we're participating, as well as the back catalog of interviews that we've done. I'm Stu Miniman, and as always, thank you for watching theCUBE. (upbeat music)