HPE Discover 2023 CTO Keynote by Fidelma Russo

I'm really excited to be here today and as the chief technology officer of hpe I have the honor of representing thousands of Engineers who are focused on our journey to transform our company from an I.T infrastructure Product Company to a leading cloud provider company but it's not about our Innovation it's about your Innovation so this morning we're going to have a lot of fun talking about how you can unlock your Innovation using the hpe Green Lake platform you'll get a good sense of our capabilities on the platform and some of the innovations that are happening on the platform you'll also get a glimpse into what's driving our roadmap for the future and hear directly from some of our customers who are using the platform for Edge for hybrid cloud and for AI and with hpe Green Lake we deliver a single unified experience across a single portfolio through a single platform to help you drive outcomes for your Enterprise more efficiently and more effectively at the edge across hybrid cloud and fully unlocking the potential of your data through artificial intelligence as we've all realized 2023 is an inflection point in the age of AI and discover 23 will be remembered as the event where hpe is embracing vendor neutrality Cloud neutrality changing the hybrid cloud game and we are playing a prominent role in AI leveraging our unique ability to build Supply and support scalable AI native architectures led by our long history of super Computing leadership and innovation so I'm here today to show you how we are doing all of this and bringing it to you on the Green Lake platform but first let's talk about why a unified platform is necessary what problem are we trying to solve so if we take a look back we used to have all of our data in data centers and colos and with the emergence of the public Cloud we thought all of our data was going to the public Cloud but it turns out only about 30 percent of our data has gone to the public Cloud started with Amazon web services moved to Microsoft Azure and now we also have gcp and other clouds but while this was going on technology was evolving and the ability to process and store data generated by Edge devices has exploded our data at the edge so now we have devices and data at the edge and infrastructure is everywhere it's in data centers it's in the cloud and it's at the edge data was first centralized it was distributed and now in this world today we're dealing with heavily distributed data and so what well what this results in is increasing operational complexity high costs and lack of control and that is what we call hybrid by accident and we need to get to hybrid by Design so in a hybrid environment you need to be able to securely connect across your data across all locations from Edge to Colo to Data Center to public cloud you need to be able to process that data wherever it is and often in real time you have to protect your data that is running your Enterprise because it is your crown jewel and you have to orchestrate your workloads across Edge locations private clouds and public clouds I'm bringing these Technologies and capabilities together with purpose and intention of reducing complexity is what we refer to as hybrid Cloud by Design now hpe Green Lake is the single platform in the industry that was designed for hybrid whether you are running your workloads at the edge in a Colo in a data center or in the public cloud it is hybrid by Design and we recognized early on the need for a unified platform so you had a single place where your services run manage and secure your distributed Enterprise last year we launched the hpe Green Lake platform on the stage we remain completely committed to a single platform strategy and this year we have been extremely busy building partnering acquiring services and technologies that Empower you to manage your hybrid Cloud environments I'm excited to dive a little bit more into some of the announcements Antonio made yesterday but before I do let's just take a look back at the fundamentals of the platform the platform at its core is a hybrid cloud services platform focused on data and AI from Edge to Cloud at the core of the platform is a set of common services such as identity and access management unified support that give you a unified experience across a rich portfolio of services from hpe and our partners starting with networking compute and Storage we provide data management services to ensure your data is secure and protected in the right place at the right time and obviously we orchestrate workloads and applications across their entire life cycle whether they are on bare metal containers or virtual machines all of that is supported by global machine learning and analytics applied across your Edge to Cloud now many times we have complex Solutions and you have complex solutions that you need to implement and for that we have world-class advisory services and managed services and as we all know digital transformation takes Financial Muscle and our financial services and our acid refresh and upcycling capabilities allow you not only to have Financial wherewithal to execute your digital transformation but also to get value out of your I.T assets that exist today and we are committed to delivering all of this from hpe and our partners so that's the platform so now that we've done a little brief run through of the platform let's do a bit more unpacking on the areas around Edge hybrid cloud and AI so let's talk about Edge we've talked about the explosion of data at the edge and now more than 50 percent of the data is created at the edge I don't think any of us would have predicted this a number of years ago businesses are more and more dependent on intelligent Edge devices to run their business so the need for secure networking capabilities is excruciatingly critical and our Edge strategy is to deliver the most agile and secure connectivity required by the Enterprise with the most advanced Cloud experience delivered as a service on the hpe Green Lake platform yesterday we announced an expansion of our award-winning hpe Aruba networking portfolio in several key areas with the acquisition of access security we've expanded our sd-wan capabilities to offer a unified secure access Services Edge sassy to meet the increasing demand for integrated networking and Security Solutions delivered as a service with our acquisition of athenet and private 5G networking we are extending our leadership in Wireless connectivity and creating one of the most complete portfolios in the market we are complementing our campus switching portfolio with the introduction of the cx-10 000 it's launched in conjunction with AMD pensando and creating new category of data center switches for a unique blend of performance scale and automation now Aruba Central is the networking application on the Green Lake platform and I'd like to invite a customer to join me so they can share their Journey on Aruba Central and the Green Lake platform please welcome Kevin director and head of digital infrastructure services at Henkel [Music] awesome thank you so the band was amazing right the band oh the bands weren't they phenomenal yeah they were can you dance no can you rock I can rock okay we'll see that later there we go good to see you so we had a great chat last week and thank you for being such a great customer but before we start can you tell us a little bit about Henkel because some people may not recognize the name yep but all of us know the stuff that they make that's right and so and a little bit about yourself yeah so uh a little bit about Hinkle Hinkle is a 144 it was uh he's uh Fitz started the program about 144 years so we've been around for a while uh but you'll probably recognize us from Loctite uh Purcell as far as laundry detergent I I know laundry detergent there you go right and then Source cough as far as beauty products so from that standpoint uh you know we're across the market uh from that standpoint we're across 460 locations uh 65 countries and from that standpoint you know we cover the globe and it's exciting it's a great place to work a little bit about myself I have been in the business business a long time as you can tell with the gray hairs but you know I came I did hyperscalers build out globally build out Global networks around the world uh have a lot of fun uh and then totally enjoy the transformation that we're in with Hinkle now it's great so we tried [Music] across your Enterprise can you talk about why you chose it and the benefits that you're getting and and then we'll get into what else you found on the Green Lake platform yeah so we started our transformation about uh three and a half years ago and we started talking and started looking about what was the what was the platform or what was available out there that could get us to a point where we had a single pane of glass and Aruba Central was really what that brought to the table so we started those conversations we started looking at our infrastructure we started looking at where do we want to go what's our Target and really what it came together was the important part was the team your team and our team us working together them listening to us what we were trying to accomplish and then it was like and then we actually worked with the Aruba Central team some of the features that you have in there today hopefully we were able to help develop those and make them even better so from that standpoint we're about 50 percent into our rollout of all new Aruba equipment and Ruby Central the ability to updates and so forth like that are amazing so we talked a little bit last week you know you've got Aruba Central on the Green Lake platform you obviously have a lot of compute yes in your infrastructure and talk to me a little bit about you know what else you found on the platform and the and the server management that we have there with our compute ops manager yeah so over the last six months we've been working with your team a lot around Green Lake and Green Lake again comes back to meeting that single pane of glass so what we're looking at being able to do now is several years ago we we basically took the position that we were moving out of managed services but not out of services that support us and help us from that standpoint so what Green Lake brings us just like Aruba Central brought us is the ability to see our entire infrastructure be able to respond to outages or actually some of it being predictive to basically say hey this is this may you may want to make me look at this this section over here and and fix that so from that standpoint Green Lake some of the functions and features that we've been seeing is definitely in aligning with our transformation in the compute and storage and especially because we're a hybrid so we have hyperscaler and we also have private cloud so why do you think it's important to all of your info times you have different management applications for those on different platforms yeah as I mentioned earlier I've been in this business a long time and it's always been this dream of mine to have the ability to look at our infrastructure at one location and utilize the tools to bring up information to be able to support our business so we have a very small team in retrospect as far as from an operation and deployment standpoint so being able to quickly identify problems and resolve them is allows my team to actually focus on the business and find Technology Solutions for the business because give you an example my lead for the Americas region the other day told me he said you know basically I had a problem I used Aruba Central and part of the Green Lake tools that we have and we were able to diagnose that problem in 30 minutes versus four to six hours that gave them opportunity to go to do other things which they like doing yeah I mean actually it's true not many people like fixing problems they're trying to focus on the future exactly so so as you look ahead what do you want to see from us so continue in that partnership uh you know and listening to us and and looking at the fact that we have a pretty spread out Enterprise we also have a pretty diverse Enterprise so I think we are a good model for you all uh and I think you guys have been amazing from the response and the listening standpoint so really just continuing to do that Grace anything else you want to share with the audience Kevin no I mean uh it's exciting to be here and it's an ingredient it's in a it's a journey and I'm glad that we're taking this journey yeah so thanks for coming with your team this week and spending so much time with us yes thank you that's great thanks Kevin take care have a good day thank you all [Applause] so as you heard Kevin's Journey started on the platform with networking but he realized he needed a unified Cloud experience and he found he could get this through hpe Green Lake let's take a closer look at what we mean by a hybrid Cloud experience so clearly we've learned that hybrid cloud is not a destination it's an experience and we have workloads everywhere and one size does not fit all and workloads in the private in the hybrid Cloud need to be placed on the right infrastructure in the right location to be of value to the customer last year we introduced hpe Green Lake for private Cloud Enterprise this is a fully managed service on hpe Green Lake we have now recognized that you want a choice of how you manage your private cloud either fully managed by hpe or self-managed by yourself so this is why we have expanded our portfolio to introduce hpe Green Lake for private Cloud Business Edition to provide a customer managed option so let's double click a little on these offers we've been busily at work on all of these options this year and we have expanded our fully managed offer private Cloud Enterprise to now allow you to manage VMS and containers across multiple clouds it provides secure zero touch provisioning capability for up to thousands of locations making it ideal for Edge deployment or for data center deployments now it's a fully managed option but you still have to deploy your isv applications on it so we have provided a click to deploy capability on the platform dramatically reducing your time to value on the platform so hpe Green Lake private Cloud Business Edition is the newest offer it is self-managed by you or you can have a partner manage it it provides a unified Cloud experience for virtual machines either on-prem or in the public cloud with a consistent experience and we start with Amazon web services and soon to be expanded to other clouds for ease of operation we provide a single click multi-system deployment capability so you don't have to go click click click thousands of times we provide a single click multi-site upgrade capability allowing for fully automated life cycle management this is all protected by our hybrid cloud data protection capabilities available on hpe Green Lake so with pcbe you get to manage it yourself it comes with a rich set of management capabilities to simplify your operations so even though you're managing it yourself it is really simple and to see these capabilities I'd like to invite Dennis and Caitlyn to the stage foreign [Music] yes so no good no good at the keynote would be any fun without a demo and I'm Caitlin here how are you feeling guys I'm so excited to be here I'm so excited for all of you to be here and I hear we're going to be talking about HP Green Lake so why don't we go ahead and dive on in that's brilliant and for any good platform the first thing we do is make sure we have the right credentials so let's log in that's important right so what do we see when we log in we see our main Control Center and as you can see here it's a bunch of easy buttons it makes things easier than ever it looks easy to me yeah that's perfect so I do hear that HP Green Lake serves two different constituencies and when I think about HPA Green Lake I hear the one that it's a Services platform so let's say I've picked out a service like one of these and I told hpe to manage it all for me but what about if I wanted to own and operate my own gear kind of like what we were hearing from Hinkle earlier today and from Antonio yesterday during his keynote how does HP Green Lake serve that constituency well as you probably will not be surprised to hear me say there's a button for that in fact some of the new acquisitions we made new technologies there are buttons being added to this system all the time so it's an easy place for you to find whatever you want to do in fact as one of the very cool updated buttons that manages your entire network perfect that's what I was hoping to see that's awesome do you want to go there should we stop firing that up you can we can share with the folks what we've been innovating on absolutely okay let's do that so obviously this is not surprising Green Lake Central so let's take a look at some of the new Innovations and visualization at Green Lake Central what we're dealing with here is a new location-specific approach to managing your hybrid Cloud distributed Enterprises right it's all there that's perfect yeah so what we're looking at here is a campus campus town could be your town it could be your location and I'll show you Caitlyn how easy it is just to Traverse this radial diagram to actually see all the individual pieces of data that this platform has to offer inclusive of security and overall alert status I see the risk is low too yes looking good you can see all the clients you can see all the traffic you can actually also see the devices as many of you know with there is Wi-Fi everywhere there's access points everywhere and it's really neat to be able to drill in in a particular device and see the information from its point of view inclusive of its connectivity so you can see the traffic through the network and everything that's exactly right and what's connected to what in case something goes bump in the night you have the ability to know where to trace the particular packets so this is useful inclusive of events when people walk through the door log on to your Wi-Fi in the hotel or a office location we capture that too an accessory definitely do you think there's any new any network administrators in the audience tonight or today there's gotta be yeah okay so if you're a network administrator you know it could be rather difficult to find out vlans and vxline and so on we have new visualization for that as well making your network easier than ever before I mean I'm not even a network admin and this is looking pretty easy to visualize and I'm sure the audience would agree with that as well because I I'm not a network admin so I need to be able to see what I'm looking at for the network configuration but networking is not the only part of it that's right we have servers networking and I actually work as an engineer in HP storage so what about storage and actually what about integrating all of it together that's a brilliant idea as you heard for demo is talking about we actually have the private Cloud Business Edition with a dude just that so part of the private cloud is of course that Computing Network and storage need to operate where the data is so therefore we have to have an easy package to roll in fire up and then we can manage our workloads does that sound like a plan oh that sounds like a great one okay so let's launch that and here we have just three little systems I'll show you some of them are reasonably idle and one is so toodling along quite nicely and we can navigate that here and we can see what VMS are running so in the case of consumption here the VM is actually the consumption perfect and what systems we're running on and it's important to understand this is not just a server this is not just storage it's just another Network it's all integrated into a location-specific system for purposes of running your workloads across your like distributed Enterprise that is what we like to see so let's take a look at what you can do you drill into one of these systems like I said VMS you have data stores we all know that could be on a sand it could be on NASA it can be local be vsan it could be anything you want right clusters what kind of clusters are you running you know this from running vcenter you have multiple clusters and also what servers these clusters are running on and finally what storage device there you go so it's all there you like storage right I love storage yes yes indeed so when you look at all this what can you do when you have multiple clusters you can let to protect each other by creating essentially a mirroring buddy and replicate bi-directionally from cluster a to Cluster B why is that a good idea protects your data and the data is indeed what keeps your business running so we've got to make sure we got that covered security is a necessity but how about let's talk about the VMS let's do that so VMS let's look at the main menu here click on VMS and you'll see here we have a bunch of VMS that we can run we can drill into them and look at the details okay great so as an owner I'm actually able to see all of my VMS on the platform right but what about the operation portion am I able to deploy a new VM of my own on the platform and is it that easy it's very very easy and that's the best part guys here is the idea we want to create a new VM let's do one should we do one right now absolutely we'll do one we'll pick it one single VM and deploy that inside the next 30 seconds how about that 30. yeah how easy it is I'm going to go through it very very quickly Caitlin attention that's perfect okay good all right so here we go we picked one we call it what a surprise discover 2023 and then we pick what kind of cluster we want to run on and we go on and as you can see this is a wizard it takes it through step by step we have the general name we have the hypervisor cluster of choice and we have the target data store let's pick that so it's like shopping right you've got a little salami a little provolone a little bread and before you know do you have a sandwich this is your Chef cooking up VMS all right how easy this is look at this how about an OS template oh yeah we need an OS let's do windows let's just go with it what do you like you like Ubuntu Santos I'm going with Windows let's do one oh my windows there's two windows I like it do any of you guys do windows all right old joke old I.T joke you know all right so let's grab Windows stick that into VM and the pause here about the data protection I like flash it's fast but I also like dedupe and compression so I don't use too much this space so I picked that policy that protects my data and optimize my use of the storage that's it let's take it here it is and now ta-da shopping list it's done and I tell the system to go do his job that is that is very easy I see that it says off though so how would you and now you can build a bunch of these and when you're good to go ahead of time you can just go in and say let's pick one yep let's go and turn them on you can actually do it ahead and go all right let me stand up 10 of these or 20 of these or 200 of those turn them on let the system know you're ready to go and bingo we're up and running it's just as easy as you said it was going to be and I do see that it says public Cloud VMS up there so is there a way to deploy a public Cloud VM and is it just as easy so so let's do that real quick let's set up a VM on the public Cloud it's as easy peasy as you saw before the wizard will take care of everything for you you click your shopping list you'll kick the instance and you're actually Off to the Races and what you're happy with in this case is C1 medium we're good to go click and Tada public Cloud your hybrid Cloud stem to turn VMS everywhere easy as you can imagine so that's it and we've all been running Isn't that cool that is awesome that is awesome wow the cloud experience everywhere all Cloud yes thank you hey guys [Music] so it's all about the hybrid cloud and it's all about the cloud experience that you get but wait we're not finished with our private Cloud portfolio changes we also introduced yesterday the hpe Green Lake private Cloud portfolio at equinix with hpe Green Lake private Cloud at equinix we pre-provision pcbe and pce at seven global locations in ready to use availability zones what does that really mean to you no more lengthy provisioning you can now get your private cloud service up and running and ready for business in days so with the announcement of our equinix partnership and the expansion of the private Cloud portfolio you now have options of fully managed or managed by you on a single platform from Edge to Cloud to Data Center however if it was only about having the right infrastructure in the right place we'd be all done but you have to operate it as well so in order to deliver a unified application stack everything underneath needs to be in sync all the way from the networking layer to the hypervisor to the workload platform to the operational complex to the Opera to the applications so in a distributed Cloud there is an increased risk of having misaligned application stacks and what does that result in poor outcomes to your business unit users as your applications are distributed across the Enterprise today most organizations have multiple teams I.T operations teams manage some parts of the infrastructure SRE operations teams manage other parts of the infrastructure they use different tools they pay more for licenses and you spend an awful lot of time integrating with Excel spreadsheets the tool of choice operational complexity is everywhere now to manage operational complexity the ability to detect Monitor and instrument are required to ensure that there's an optimized application stack for your business you need a common operating experience which is agile provides real-time insights to get accurate reporting reducing your costs accelerating remediation and all of those are now major requirements in the hybrid Cloud world I spend some time myself managing I.T

and I know how complex it is it is one of the toughest roles in an organization and the critical role of itox and SRE teams is fundamental in realizing the promise of the hybrid cloud so we realize that you need a single operating console that gives you a consistent view across your distributed Cloud a console that allows you to instrument and get visibility at every layer of the stack the ability to analyze the data so you can quickly remediate it you also want a console that is multi-tenant so you can use a single operational tool across multiple teams and make sure that they have guard rails you also want a tool that is multi-tier so you can consume the service yourself or you can use a service provided by an MSP or a partner and this is why we acquired Ops ramp which provides the full complement of operational tools to manage the complexity of today's hybrid cloud using the power of Ops ramp on the hpe Green Lake platform you can automate device Discovery you can gain a unified view with full stack observability you can eliminate repetitive tasks and you can accelerate issue detection and resolution this is all on a single platform to use across all of your teams and it doesn't matter what resources they are and from what vendor because we are committed to being vendor neutral Cloud neutral in a multi-tenant tool provided in multi-tiers so you can consume it whatever way you want and with optramp we give you choice and we reduce complexity in the hybrid cloud and in addition to vendor neutrality and Cloud neutrality we have more than 2500 Integrations to empower you to discover Monitor and remediate your idea State and with that I want to bring up one of our Ops around customers to talk about the importance that of having visibility about across their entire multi-vendor estate please welcome Michael director of it monitoring and automation at echostar [Music] hey Michael good morning good to see you good to see you again I think I saw you the other night we did see the other night thank you thank you thank you so tell us a little bit about yourself and about icostar before we start all right Echo stores uh global satellite and Communications company we provide Communications and additional services to our customers my role I'm a director of I.T automation my team is responsible for I.T operations and management throughout from the physical to the cloud so you're the guy who keeps the satellites up and running no not satellites I.T stuff I.T stuff I.T stuff but that keeps the satellites up that it does there you go so so can you uh when we last spoke uh you talked about your journey looking for an I.T

operations tool and I think you you know went through a long evaluation cycle can you just go through that Journey for for the audience and share you know the cycle and what you did sure can so we started looking for replacement of our existing monitoring tools because we were growing too big and we started about a year year and a half ago and we evaluated and did a handful of uh proof of concepts with multiple vendors Ops ramp being one of them and we went through down to details on every single product to make sure we met had something that met all of our needs so do you have multiple clouds do you have on-prem can you share a little bit about your environment we have a very complex environment we have grown lately like everybody else we have a large on-prem facility and we also have something in every single Cloud that's out there so so you chose evaluate two and then we get to actually stalling them and having them run there's there's often a disappointment between the expectations that we had and and how and you know what our experiences you share a little bit about that uh with us absolutely so we had Opera we were in the process of implementing and we were struggling sometimes with things didn't work the exact way we thought they would or the thought way that we thought they should we actually ended up having developer calls with the Ops ramp developers in India and my team and they either developed or changed how it worked to meet the needs that we had or they helped us change our Paradigm on how we look at the systems that we're monitoring so you have a very interesting story here here has asked both agent agent capabilities and when we spoke last week you mentioned that you actually chose the agent route and I uh you know come from the school of you know agents were around agents were bad now we launched an agent list and now we're back at agents again can you share for the audience a little bit about your thought process and what your experience has been sure so we were actually moving from a mixed agent and agentless solution as it was and the experience we saw was when we had an agentless environment in the middle of the night we're doing backups and we have some systems that we have to prioritize the backups to get it off the system and that flooded the network which then made the agent list monitoring tools not be able to connect and we got a lot of alerts and that created alerts in the middle of the night often and so we decided for servers we're going to an agent-based solution we're going to enable it so it has a priority to send out its its status on a regular basis and we were actually surprised at the quality of the information we got by having the agent on the system so did it take a long time to roll out it actually did not it only took about a month to do all of our rollouts and now your calls have been like can you tell us like what what's the impact been on the number of calls so the number of alerts so I'll talk about the calls too but the number of alerts went from um the normal several hundred a night down we dropped by 80 percent just by having the Ops ramp Tool uh correlate the different issues that were happening so everybody is less cranky in the morning especially me that's good that's good we want you to stay that way so so when you heard that hpe acquired auxram what was your reaction my first reaction was not a very happy one I really liked the way operant was I like talking to the team that we had they were really responsive to everything and then as I started looking at what HP was doing I've changed because you're not just buying it and putting it inside you're actually incorporating it and making it better you know having additional staff that they can actually use for the AI and grow their skill sets and their their abilities really is a good thing in my mind and what do you think about it being on the green light platform uh from my experience on the Green Lake platform it is phenomenal it is it's going to just be more stable more available and more capable as we go great great so is there anything else you want to share and we and we are committed to making this uptrend acquisition you know kind of the center of our operational capabilities uh for the platform and delivering more and more as we go forward here anything else you want to share with us or with the with the team or the audience here this morning the biggest thing is their people are awesome I love how they've jumped to help us with everything um they're they have people around the world like HP does HP does now and they jump when we have stuff in the middle of the night we have people that can call my team has relationships with the developers that are phenomenal and it's been awesome that's Grace that's great well thank you Michael and enjoy the rest of the rest show and I look forward to working with you in the future absolutely thank you all right thank you so Michael's Journey started as you heard looking for the best item platform on the market and since the acquisition closed we've been busy integrating it into the hpe Green Lake platform so I would like to reiterate how Ops ramp is transforming the operational capabilities that we provide you for the hybrid cloud with the integration into the platform the hpe Green Lake platform is now vendor neutral Cloud neutral multi-tenant and multi-tier and AI driven but to take a good insight into that let's get Dennis and Fatima on the stage to look at the journey and experience that Michael just talked about okay we're back we're back but you thought you're gonna get by so easily with only one demo no no how are you doing Fatima I'm good how is everybody [Music] so Dennis I'm gonna ask you a couple of questions today are you ready I'm ready do your worst awesome okay all right it's dead and so we saw fidelma talk about having instrumentation at each layer of the stack right and we have seen that the it ecosystem has changed so much it has become much more complex with having your data and services and workloads across all of those different locations Edge data center public Cloud private cloud so tell me Dennis how does hpe Green Lake help with having that visibility across the stack how does it help me navigate that complexity well you know what I'm going to say we have a Green Lake and on Green Lake there's a button for that oh yeah oh there we are this is actually our good friends up upshram right Upstream got purchased by us you know that and what offram is doing they have 2500 Integrations and they are truly 100 Cloud neutral and vendor neutral awesome and I see there it's Dell and NetApp is that for real that's for real awesome that's for real because if you can't see the problem no matter who the vendor was you can't fix the problem so we instrument everything in the stack and we all know as much as we would love you to buy nothing but hpe that may not be the case right you may be making an acquisition and they had Dell and super micro and so on Everybody Plays in this world and therefore it has to be able to be vendor neutral awesome tell me more about that oh there's all kinds of stuff here we can go through this for the last 20 minutes but we don't have 20 minutes so but I want you to see that there are things both in the public Cloud that can be instrumented that's part of your stack in the public cloud and then there's also part of the stack that's running on premises okay so you begin to have this hybrid Cloud Dynamic and you instrument both sides of the equation and you said it absolutely right Fatima it's like you might have a hundred locations a thousand locations depending on how geodes first you are so those locations need to be instrumented and when we do that it has to be hey what if you have Cisco out in one particular occasion what if you have another vendor in another location yeah whatever I do yes it has to be it has to have the ability to openness at all levels exactly so um so now with this in mind though can you pick from a large menu and turns out there's actually a huge menu to pick from several hundred service snap-ins several hundred middleware snap-ins all kinds of stuff switches Galore you get it you pick from this nice menu and uh you'll be picking for a while that's fantastic yeah so what can you do with it that's the key okay how do you visualize it one of the biggest problems here is complexity right how do we actually get complexity under control so we don't choke on it and the way to do that like I said if we instrument things we can see things yes and how do we see it easily quickly in a simplified way so we've talked just the other session we talked in the other demo about VMS so maybe what we should do is pick some VMS how about pick an address for vcenter a lot of you are using vsender right yep let's do it let's do it so what visualization can we see when we know oh here's a v sender running a cluster or two what does that look like and it looks rather simple doesn't it it does so I mean my environment is going to be a little bit more complex than that so can we go deeper yes we could go deeper because we have the ability to not only examine the individual objects in a map like this but also sneak in and set the aperture of how deep we want to go so we want to go from three to a depth of five we actually begin to see a different picture are you ready I am ready all right let's take a look how about that ladies and gentlemen a beautiful drawing of a two node vsphere cluster that is so elaborate it's really cool I mean we don't have time to go into all elements here but you can see there's local data stores there's a couple of VMS and there's two NFL shares in the middle pretty nice and here again we can drill down see the status of things the IP address and in this case is running a Windows Server um everything is fine everything is up and running so now you can visualize what's going on you actually have a fighting chance of being able to manage it yeah I I think that's pretty awesome because then that will get me to a faster root cause analysis because I know now what is connected to what that's exactly right and if you need a little status of all your estate guess what we actually have all the VMS roll up into a graph like this a dashboard that shows you the health of the virtual machines running in your shop yeah and I see some performance metrics in there as well so that's pretty awesome so you got a pretty comprehensive view I see yeah so a lot of people are saying well that's fine but I'm actually not just running VMS I'm running things like what about kubernetes you mentioned yeah so how about that you can actually go up the stack and say I don't care about all the hardware I don't care about the the virtual machines I care about the kubernetes map what does that look like and here's an example of that diagram and this is pretty wow this is too deep it's pretty powerful stuff I mean it is it is really fun to play with this tool and I encourage you all to do it there's of course a dashboard no matter where in the stack remember fidelma talked about a stack you can drill in at all stack levels draw your Maps see the dependencies and so on and it's only for one good reason right we are going to see if we can fix things yes so that's where I would like to ask my next question which is about remediation so I love that we can see and view all of that and we can instrument all of that but if I'm an I.T decision maker I would like to know how we can get to remediation because I care about my slas I want to make sure that all of my services are up and running smoothly and if there is an incident that hits how can I quickly get it up and fix it let's take an example what do you say let's do that so here's an exam you can see the customer portal is down the Enterprise Erp system is running the building system is running that's all fine but we're getting a red alert here on the customer portal so it'd be nice to be able to drill down and see the relationships of all these processes running across the entire ecosystem and this is what we're mapping here that's good Isn't that cool yeah if you can see the problem I mean I can I can see how applications are connected that's pretty cool and you can see in the middle everything is fine on the sides right yeah but you see that that there's a there's a bandit in the house here we had my sequel cluster is sitting there not doing too well as you can see when we examine this we're finding it's been down for several days wow I wonder what's going on and this is where Auto Reading mediation comes in okay tell me more yeah so let me show you one of the coolest things about this system it actually uses AI if you have an alerting and alarming systems before you know you can drown in alerts yeah there's a lot of alarm fatigue there yeah what do you learn to do quickly you go I'm going nuts I'm turning them off well now you're Flying Blind that's not so good right yep so here how do we actually normalize because you got to remember this layer cake not stupid to be 100 aligned when it's misaligned what happens is you get alerts from all the different layers from their vantage point so what we do is we normalize this into a data stream and that data stream is fed into an auto pattern recognition AI engine that has seized patterns before and gets smarter and smarter all the time and it correlates everything to the net event that is the root cause of all the alerts you're getting so I only see actionable data and events that's pretty cool think about think about a fourth and you we heard it here right is that it turns out that I used to get hundreds of alerts every hour now it's down to thus the essence I can act upon that's where you pass and I know you guys have been into business for a little while you know we always said we wanted autonomous systems systems that can heal themselves well now it's upon us let me show you real quick you actually set up a flow diagram to Auto remediate once you understand the root cause and when you do that ta-da you get everything up and running again this is awesome Dennis so what I'm hearing is you can not only instrument you can view and then you can alter immediate so that's super powerful thank you so much for thank you so much all right thank you guys thank you guys great job guys so as Dennis and Fatima said it's all about AI and 2023 is the age of AI it's the beginning of the age of AI so hpe is a world leader in super Computing and Enterprise grade Mission critical architectures all of which are key elements for IAAI so we recognized early on how important scalability is in order to achieve the promise that AI can deliver and for the vast number of years we've been making strategic Investments both organically and inorganically in AI Technologies and high performance Computing to build a foundation for the AI explosion and its impact on how we live and work take a look at some of these Investments and how they have uniquely set the stage to empower your AI Journey the acquisition of cray demonstrated early on our investment in leadership and high performance architectures the development and delivery of the frontier supercomputer and large-scale application support further Advanced our journey to deliver ai-centric architectures to run the world's large language models however having the hardware architecture and system architecture is not sufficient to bring the power of this architecture to llms and generative AI and to that end we have invested in an AI software platform and created an end-to-end platform that will enable you to train and optimize your llms our acquisition of determined AI gives you the ability to train these very large models at scale or to use the model Hub to pick from existing large language models our acquisition of Pachyderm gives you the software you need to automate repeatable or reproducible machine learning data pipelines for large language AI applications and finally our Enterprise grade proliant and high performance Computing portfolios deliver the ideal platforms for you to run inference on these AI models as Antonio announced yesterday we now have hpe Green Lake for large language models as a service on the Green Lake platform so let's unpack what hpe Green Lake for large language models offers first hpe Green Lake for large language models is an is service it's built on the world's most sophisticated super Computing technology and it allows you to train optimize fine-tune AI models at scale in any location for Enterprises of any size making it accessible to everyone I think of it pretty simply instead of having to buy a supercomputer of which I'm sure many of us do not have the budgets you can now simply subscribe to hpe Green Lake for large language models as a service and there you will have access to one of the most powerful supercomputers in the world to train your data or leverage our model Hub where you will have models that are developed and trained by yourself or other people and there will be a button for that on the Green Lake platform making AI accessible for all now ai is top of Mind strategically for all of our CEOs and in the boardroom but there is another conversation that is intrinsically tied to AI which is sustainability which is also on the top of Mind of our CEOs and boardroom members and as we think about sustainability in the context of AI how do we manage the carbon footprint of training tuning and running these AI models how much power does it take to train large language models and generative AI so let me share some interesting facts take for example gpt3 it is the precursor of chat GPT and training gpt3 use the energy equivalent to 30 average U.S homes for an entire year or as someone told me the other night as much energy as the Las Vegas Strip in one day the latest models have doubled if not tripled that power consumption and that is only for training the model it turns out that the power consumed to train the model is about 10 percent of the power budget and the power consumed during the inference phase is at least 9 to 10 x more than used to train the models and this is why High availability designs are critical and low power designs are critical in driving the costs and power requirements of these models empowering our hpe grain Lake large language model offer we take this incredibly seriously and we are Engineering in sustainability end to end our areas of innovation include liquid cooling of which we are an expert power supply design server design and overall High availability system design so models complete faster for both the training and the inference platforms we continue to improve our machine learning development environment to optimize the models to allow you to train the models faster and with less gpus so you can reduce your power consumption by up to 20 percent hpe Green Lake for llms runs in 100 percent carbon neutral co-location facilities and the heat generated is used by the communities to power different initiatives such as greenhouses Green Lake Cloud platform is one of the most sustainable platforms on which to run your large language models now because hpe Green Lake for llms is running on the hpe Green Lake platform you're able to use the sustainability dashboard to gain insights into the energy consumption associated with running your AI models across the hybrid cloud let's take a closer look at the sustainability dashboard which was officially launched on the Green Lake platform on June 5th so we have been developing the sustainability dashboard to help you monitor observe and give you insights to help you reduce your energy consumption across your entire it estate the sustainability dashboard it delivers key insights on it energy consumption carbon emissions and electricity costs based on location and we take in analytics from Ops ramp and the HP portfolio for compute storage and networking so we provide insights across multi-vendor and multi-cloud it Estates and I'm also thrilled to announce that we have multiple customers such as Nokia danfoss suiko and the Ryder Cup actively evaluating the dashboard on the hpe Green Lake platform to give you some more insights into the sustainability dashboard I'd like to invite Dennis and John back onto the stage [Music] hey guys hi everyone now we're going to talk sustainability yeah Dennis do you got one more dashboard for me we got one more dashboard this is someone like cherry on top of a three-course meal today because of course you know now there's going to be a button for that there is going to be a button for that and Dennis like fidelma just highlighted sustainability is imperative as it organizations across the globe are leaning into a more sustainable way of operating there's a lot of new challenges that they're faced with what does hpe Green Lake offer those organizations to help keep their sustainability in mind yeah again you know John I'm going to say if you can problem you can't fix the problems definitely this is about using Green Lake to make sure we're more sustainable and indeed we have a button let's click that button so let's go and take a look at how we look how we instrument and actually Bubble Up using Ops frame the data we get from all those devices bubble into one easy to read overview across your entire estate here you have it John wow carbon emissions right we have we have energy consumption and we have energy cost wow those are the three big uh pieces of data that we have to keep in mind when talking about sustainability that's right and now comes the sexy part right where is all this data coming from like I mentioned it bubbles up through the instrumentation of everything in your estate into a long long list of stuff that is all aggregated together from one large data Lake if you take a look here John this is really cool of course you would expect okay HP probably has HP gear instrumented yeah I think you can count on that right and in fact we actually here see some Dell servers too wow so it's not just hpe only technology we can have open access to any other vendor on the market 100 we can again because of Ops Rams brilliant instrumentation thanks to office Ram right so here is the where you see the roll up and of course what you can do is begin to filter things yeah I know some workloads tend to be a little burstier have moments of higher traffic uh can we try and take a look at one of those time periods and see the sustainability metrics of our workloads exactly John because what you end up with is with that much data you have to be able to narrow the scope of what you're looking at so let's start with a simple thing a nice little calendar where you can go in and pick time and I'm going to go through this pretty quickly but you'll notice as I pick different dates can you see the graph shows a different roll up yeah everything changes just when you click on it yeah Isn't that cool school so the ability to to real-time query this gigantic data lake is actually very valuable you can do what if analysis take a quick look at what's going on what happened last Friday and so forth for example we could do the month of May John let's do the month of May let's say the month of May what happened in the month of May here's the entire month of May rolled up across all devices but that's not all that's not all what else can we do to slice and ice this data well we can actually filter as you would imagine right awesome so let's put a little filter in right it's like a coffee filter we don't want the grounds we just want the good Mo good job we just want the good stuff yeah exactly so let's kick in a filter here of course this is super important to understand guys you're actually dealing with objects that have attributes and we can filter on the attributes and in this case let's have some fun and filter on the model attribute what do you think of that let's do that that sounds good okay so since we talked about the fact that we can see other guys's stuff too let's just pick the power the power profile of our friends from Adele let's do it apply that to the I know there you have it how about that wow that's super awesome to be able to take a fine-grained look at different pieces within your it estate to see their sustainability match that's right that's amazing it's like a big pizza pie where you say hey you know what I don't want a whole Wagon Wheel I just want to slice you can slice and dice the data any way you want yeah so Dennis I see we have three awesome dashboards right here what if I wanted to do some of my own calculations run the numbers on my own is there any way that I can create custom reporting to make things even more actionable well I'm glad you mentioned that because we all know it doesn't matter how great your reporting tools are your boss will always say yeah but I want to see the data this way right so how do you do that well here's the good news to have all this powerful data and one easy to place shop where you could just extract it we can export the technology we can export the whole database if you want and let me just clear the filters John pop up to the export button let's export it and let's get it out of the system so you can have at it with Excel and Report generators whatever you want yeah so here we do it we just export and most simplest thing we can do comma delimited CSV file export do your thing Mr little system let's do it let's do it so here it is generating a report look at how easy that is precisely right it's all about the data so you are powerful you can do what you need to do across your entire State thank you very much guys appreciate it hey guys thank you thank you so as you've heard we take sustainability very very seriously it's core to our foundational strategy so it's been quite the morning we've talked about how the Green Lake platform can Empower you to run your business and I've shared with you many of the exciting Innovations now available on the platform across Edge hybrid Cloud Ai and sustainability but to bring it all together in the voice of a customer let me bring in my next guest Jeremy senior director infrastructure in operations at trinquero thank you thank you welcome thank you so thanks thanks for joining this morning and um you know trinquero is the second largest family-owned Winery in the world and many of people here may recognize the brand um but not recognize the name so it's uh it must be a great honor to work there so can you tell us a little bit about yourself and your mission entering Carol absolutely well first of all thank you for having me this is a great honor uh to have us up here today and uh trinquero has a 75-year history we're celebrating our 75th year and uh for it's founded in 1947 for the first 30 or so years we had a single brand and that was Sutter Home some of you made very familiar hopefully and uh and then we went through an expansion phase in the 70s and 80s uh acquired a couple other properties uh went for another 20 years and then had a rapid expansion in the in the 2000s and I came in in 2016. as senior director of infrastructure and operations I'm responsible for the infrastructure Network cyber security web development database development and it's been a it's been a fun Journey so why did you join oh I was looking for a new challenge at the time and and uh trinquero had that challenge they'd had this rapid expansion technology uh had not quite caught up with the expansion so we needed to modernize technology we had some uh processes around that technology that needed modernization and really just needed some new thinking and so uh it was a it was exciting and uh I went for it so you told me something a funny story the other night about you came in but it was all solid for you it was all solved all solved you know exactly yeah there was a there was a consultant that had been in and and the the conclusion drawn by the consultant and this was very popular in 2016 was well you just need to move to the cloud if you move t

2023-06-25 21:42

Other news