 From theCUBE Studios in Palo Alto in Boston, connecting with thought leaders all around the world, this is a CUBE Conversation. Hi, I'm Stu Miniman and welcome to a special CUBE Conversation. Digging into some of the hottest topics in tech, of course, multi-cloud has been one of the big things we've been talking about for a number of years, the maturation from just cloud in general, hybrid cloud and multi-cloud. Happy to welcome back to the program, one of our CUBE alumni, Chandomai Mondal. He is the director of marketing from Dell Technologies. Chandomai, pleasure to see you. Happy to be here. All right, so last year we were together for Dell Technologies World, VMworld and of course, I've seen how these solutions have been expanding out, partnerships, especially a lot of it from Dell side, leveraging VMware technologies to extend and connect to what your customers are doing with their cloud strategies. So give us the update as to what you're hearing from customers and how Dell is moving to meet them. Sure, cloud adoption is really growing. And even from the three hyperscalers, AWS, Azure and Google Cloud, there are over 500 different surfaces today. And with this past phase of innovation, I see customers adopting many different surfaces from these public cloud vendors. And again, they want to adopt these surfaces because they are differentiated. They have workloads that can leverage the surfaces and sometimes even from leveraging the same dataset. One challenge that we are seeing is how do customers move data around from one cloud to another so that they can take advantage of the great innovation that is happening with cloud storage or cloud provider because moving the data comes with not only the migration risk, but also huge egress fees, the time it takes. So solving this customer challenge is our number one priority in the cloud offering. Great, Chenamai, you brought up a bunch of really good points there. Of course, nobody solved the speed of light issue. So we know data has gravity, it's not easy to move it. And yeah, absolutely, I've been saying for the last couple of years that data is one of those flywheels when it comes to the cloud. But once you've got it in there, it's not kind of the traditional lock-in, but I have access to the data. I have access to the services and it's not easy to move it out even if customers would want to take advantage of multiple services from multiple clouds. So I'd love to hear, what's Dell's role in this discussion? How are you helping us make our data more of a multi-cloud enabled environment? Absolutely, absolutely, Stu. So with us, Dell Technologies Cloud Storage for Multi-Cloud, we are delivering scalable, resilient, cloud-attached storage with flexible multi-cloud access options, ideal for securely deploying or moving demanding applications to the cloud for many different use cases. The way we are doing it, effectively, that customers can leverage, block, or file storage consumed as a service directly from any different clouds like AWS, Google Cloud, Azure, and we are providing very high-speed, low-latency connections from Dell EMC storage from a managed service provider location using the Direct Cloud Connect option. And let me give you an example. We have Dell EMC Isoland Industries, number one, Gelout NAS. It has very high performance, drives large throughput, scales to multi-petabyte use cases, have multiple different protocol access simultaneously from many different applications. Now, the same Isoland today can be leveraged as Dell Technologies Cloud Storage with direct access to Azure, Google Cloud, AWS, consumed in the cloud operating model. So now, you can run your applications in any cloud while having data sitting outside of your cloud with the high-performance, high-speed access that you need. That's where we are bringing the innovation and the value. Okay, and if I heard you right-hand of my, this is a managed service solution because if I want that high-speed direct connection with Azure and with AWS, normally I need to be in some service provider. Dell, of course, has lots of partners that offer those services. I'm not just talking about connecting my array that I have in my data center connected over the internet because that wouldn't necessarily give me the bandwidth and performance that I need. Did I get that correct? Yeah, absolutely. Because again, you need this connection and co-location with the hyperscalers to get the high-speed connection, say in the case of Microsoft Azure, the Express Route, you need to be co-located in a facility like right next to them so that you have the high bandwidth, high performance that you need for this application. Yeah, that makes a lot of sense. It's kind of your hyperscale adjacent. You just throw that connection. It's relatively close. Might help if maybe if you've got a customer, an industry example of what the real life expectation and use cases for a solution like this. Sure, so let me give you the example of our genomic analysis. Now, is genome sequencer in a single cycle for a human being that creates 100 gigabytes of data and that's just like raw data? You need to run analysis, different types of analysis to the effects that are drug or something having on the DNA. Now, for example, NVIDIA Parabrix is a popular sequencing software that needs to be run on this dataset. And again, it drives very high throughput. Sometimes it needs 100 gigabytes per second throughput to drive the performance. Now, we have worked with Microsoft Azure very closely in using Microsoft Express Route. You can actually get that bandwidth that throughput for running Parabrix or next gen sequencing VMs in Azure leveraging ISA learn. And in fact, we have worked with Azure to provide a completely aggressive data movement. So when this application is writing back data to this application array, to the array as part of the technology cloud storage, there is zero fee associated with it. And it's not just Microsoft Azure, right? You can have the same dataset and run this Parabrix or next gen sequencing VMs in Google Cloud, AWS, Azure simultaneously, thereby scaling up this process much faster. So if you are a pharmaceutical company trying to cure for a disease spreading across the globe, you need to run this on hundreds of thousands of patients creating hundreds of terabytes to petabytes of data. Then you can actually scale up the process across three or more different clouds very quickly. This truly shows you how you can leverage the power of ISA learn, the scalable high-performance storage in a multi-cloud world. Yeah, very interesting. You talked about no cost for egress fee and that can be one of those architectural killers. You think you have a good solution for a cloud, you put things out and then all of a sudden you start getting things on your bill that you weren't expecting. Is there something special that the customer needs to do for this service? Is that you're saying is that a partnership with Dell and Microsoft? Or how does this differ from the traditional egress fees that I'm used to getting for whether I'm using AWS, Azure or Google? So this is like a Dell and Microsoft Azure partnership. So that's where you do not get charged with the egress fee when the applications running in Azure connecting back to LEMC storage as part of the cloud storage services. Okay, excellent. Because yeah, I mean, Chenemi, I'm sure you're well familiar. A lot of times people look at cloud and they're said, okay, when I look at the economic, if it's compute intensive, it makes a little bit more sense. If it's data intensive, there's lots of reasons that it might not make sense. So this is unlocking some of that data capabilities. I guess that leads to some of the opportunity around AI is of course, I need to think about my architecture. A lot of times data is not going to leave the edge environments, autonomous vehicles, an obvious use cape that we talk about. Usually there's training in a central location, but then I need to be able to actually do the work at the edge. So what does this cloud storage for multi-cloud, how does AI fit into this old gut? So yes, for AI, you need to train very large data sets for a long time and to get to the results like you opt in or you want. You gave the example of autonomous driving, right? The self-driving car need to understand many different scenarios, whether it's an IC road, a kid on the road, it's a slippery condition, or they are running into a brick wall, so on and so forth. Now, when it comes to dealing with this petabytes worth of data set, and you need to train these models, okay? You need very specific servers, GPU powered servers, okay? Now, to scale, you think that you'd go to the cloud and then you will be able to get the computer need. However, turns out cloud is not an amorphous homogeneous place. Between the vendors, there is huge difference in terms of what GPU powered server you can get. And even within one particular cloud vendor, depending on the region, this vary widely. So it becomes critical that you can have data set that can be connected from many different clouds, from many different regions as you need it. And one more thing I want to highlight, AI is actually one area where these cloud providers are providing very differentiated services. So in the autonomous vehicle example, there are several stages of model training depending on like what you are trying to achieve at one point in time. Now, you can one day, or for one part of the process, you can leverage AWS PageMaker for your model training. On the other part, you probably would like a TensorFlow for from Google Cloud to do it. Now, when you have your data set outside of the cloud and you have the fast connection from many different cloud, you can take advantage not only the different GPU powered servers, but also differentiated software services that are available from these cloud providers. All right, so, Chandamai, how does the VMware cloud solution fit into this discussion? I know it's been a important piece of the Dell Technologies cloud piece. So how does the multi-cloud storage, VMware cloud and the multi-cloud piece fit together? Sure, so VMware cloud on AWS is one of the key offering that we have, and it also fits into the multi-cloud story very well. Actually, let me explain that with a customer example. We have one of the world's largest energy company down in Texas. They have four petabyte data lake on Isola. And this is all seismic data. They are running analytic workloads to figure out exactly which place in the ocean they should drill, and precision on here can be like millions of dollars of different. Now, they wanted to set up a secondary data center in the case of a disaster. What we were able to do is to spin up a PR service for this customer leveraging Dell Technologies cloud storage. So they replicate the data to the cloud, and then we spin up their PR environment with VMware cloud on AWS, okay? And now the data is already in the cloud. So they got their PR service with VMware cloud on AWS, but with the same dataset. Now they are running those seismic analytics workloads from AWS, Google cloud, and Azure. They are by speeding up the process of finding the next location to drill. So you see the example where we leveraged DMC on AWS for DRS as a service, and since the dataset is already there, now they are running their analytics workloads for their regular operation. Great, well, definitely a quite a bit of maturation in the Dell cloud solution, how that fits into multi-cloud. Help put a point on it, Chandumai, if you would. The conversations you're having with customers and Dell's role in the multi-cloud discussion. Sure, so there are two important things. First, the ability to scale to many different clouds, to leverage the different surfaces, the compute infrastructure, so on and so forth. And the second part of it is, depending on the applications, right, you might need to leverage for the same workload working on the same dataset, different surfaces from different providers. Dell technology's cloud storage for multi-cloud is enabling that for our entire customer set. And I will close out with one more important aspect. If you are the customer who is just starting your cloud journey, or one single cloud provider calls your cloud needs for today. But still, you want to architect your solution so that when the need comes, you can actually leverage multi-cloud for compute or other surfaces. So if you decouple your surfaces from where your data is, while doing the cloud access, that actually makes your cloud architecture future-proof. So with Dell technology's cloud storage for multi-cloud, we are helping customers not only today, but also for future. All right, well, Chandramayandal, thanks so much for the updates. Congratulations to the team on the process. And look forward to talking to you again in the future. Thank you. All right, I'm Stu Miniman. Thank you so much for watching theCUBE.