00:04
next up I'd like to introduce uh Vamsi uh who is the group product manager from Azure Storage Vamsi is gonna come out and he's gonna talk to us a lot more about our integration with Azure
00:14
around the Azure native service and Azure Vmware uh service Vamsi thank you Rob uh we're delighted to be here my name is vamsi uh
00:33
product manager in the Azure Storage team been on Azure for 18 years from when it was a 25 person research team and was called Project Red Dog Project Red Dog not even called Azure
00:47
and it's been fun it's been a great journey to be on this like seismic shift you know from on premises into cloud something that's still kind of
00:56
rippling through our industry but you know to be able to be at this other size mix shift even more size mix shift to AI has also been a really fun journey
01:09
um and I'm sure this is on all of your minds as you know probably don't go through a single day without hearing the word AI at least 10 times
01:16
if not more and kind of figuring out what do I do with this thing like I'm a storage person why is storage you know and data you know relevant
01:24
what's the relevancy what do I do so I really think there's a lot that's happening there particularly around getting your data to the cloud um you know this is a quote from our Chief Partner Officer
01:38
but you know really the the key here is why Azure why pure right I'll start with why Azure um and we've been very privileged to have sort of
01:49
a front row seat at this um Azure is the cloud for AI you know the biggest foundation models in the world whether they're from open AI or from Microsoft's own AI have been trained on Azure supercomputers
02:03
uh for the geeks and all of us if you go search for Azure supercomputer Fairwater there's a great blog post that went up last week talking about our newest uh data center uh
02:14
1.2 million square feet of data center space a single supercomputer coming up that's come up in Wisconsin um the storage system for that data center for that supercomputer is five football fields long
02:30
that's how many racks of storage we have in there to feed the GPUs the hundreds of thousands of GPUs from Nvidia they're just sitting in this liquid cooled data center building next to it so
02:41
that gives you a sense of the scale that is needed to build and train these foundation models um and in addition to that we provide the broadest set of models
02:53
with the model integrations that we bring so you know whatever model you wanna use whether it's one of these big foundation models or you wanna have domain specific
03:03
custom trained or fine tune models you can really integrate them very easily in Azure and build your applications on them and the last piece of this is the tools to build these applications right
03:16
with a copilot studio and AI foundry we really provide what we think are the best in class tools to go build AI applications but all of these things come with one thing
03:30
to really unlock AI for your enterprise you have to bring your data to those applications right without that grounding knowledge about your applications you know AI is not really that useful
03:46
and I think that's where pure comes in because you are running your enterprise applications and your enterprise mission critical data on pure today and
03:55
and pure has really invested in making sort of cloud a destination for block storage and providing that seamless uh you know translation you heard Charlie talk about
04:07
you know the transition to the enterprise data cloud and we've really been on this journey together with pure for over six years now we've been working on different things kind of engineering team to engineering team
04:18
really building foundations from the ground up we in you know released Cloud Block Store a few years ago customers are using it we have a bunch of interesting success cases
04:29
I'll just cite a couple of CPG customer major CPG customer you recognize the name of immediately you know use this pure uh um block
04:37
cloud block store to get a 12 is to 1 data GDO production uh over 40% savings on their TCO uh there's a company called Perdoquia Education that really does all of their Dr
04:48
for their mission critical databases in Azure and you know cut down their restore time from 8 hours to 5 minutes uh and another major manufacturers again has been able to look at 50% savings
05:01
and reinvest that savings and the time in AI so pure has done a great job of like bridging the on premises and cloud pieces and you know Rob talked a little bit about the Pure
05:14
Storage Cloud and the Azure native version about that so let's dive into that and look at what that is and what does it mean to be Azure native right it means that the service looks and feels and operates from an automation perspective
05:29
or a manual deployment perspective just like an Azure service it's really seamless you can kind of go do all of this from the Azure portal or from your scripting and automation
05:39
and your template deployment templates uh it has the security trust relationships with the other Azure services so you can really interoperate uh cleanly
05:48
uh and you know that that whole experience is really really clean that's cool we can ship services
05:54
but what's the business value for you so what we decided to do was as we ship the service you know focus on one core scenario and really nail it and that scenario is Vmware migrations into the cloud right so what we looked at was we
06:09
have a service called Azure Vmware Services that allows you to seamlessly migrate Vmware instances into Azure and still retain all of that expertise the automation everything
06:21
that you've built up around BMR over the years you know be able to bring all of that into the cloud you know one of the way we kind of plugged all these three things together is to really make pure storage Cloud
06:32
a really really seamless experience of seamless migration better resource utilization from the perspective of you know employing and using the dynamic scaling in the cloud
06:42
maintaining all those enterprise capabilities and all of this is fully managed you don't have to deploy anything you really don't have to kind of struggle with getting environment set up
06:52
so rather than kind of keep talking about this I wanna you know show you a demo um so we're not tempting the demo gods this is recorded but I'll talk over it
07:01
ha ha ha but um let's look at this and so what you're seeing here is the Azure Portal and you can see this you know typing in and looking for pure storage
07:10
shows up just like another Azure service below that are the Azure native services Click create provide some very basic information to create an instance of the Pure Storage Cloud in your uh deployment
07:24
quick couple of clicks create this creates an empty shell that you can now go create a storage pool in and then create your volumes in so that's kind of where you know
07:34
you see this new thing you could go and click on that instance and say well I'm ready to create a storage pool that's where we do all the storage stuff that we're all familiar with
07:43
and you can see here again looks completely native because it is right you're selecting your availability zone you're picking how much throughput you want and how many IOPS you want
07:53
your storage is build pay go so how much ever storage you use is what you get build for networking config and you're deploying right this takes a couple of minutes
08:03
so we'll skip forward and you have a storage pool now let's switch over to you know and you go into this resource it's you know kind of empty nothing's really happening here
08:12
let's go really create that storage pool in in VM Sphere and and pure one other additional step that's happening here is you're connecting it to your Azure
08:24
Vmware solution that was a couple of clicks as you saw there now here you are in this familiar interface with Pure Storage and Vmware you've created a um
08:34
data store that that shows up for you you're able to kind of quickly go ahead and configure that give it a name and a size and all that and you know no default policies
08:47
but that creates something that your vsphere client can actually see and now from here on it looks just like all of the Vmware work you've done before creating a new VM called Demo VM
09:00
and you're really selecting where your storage is so storage pool data store that we created earlier and you know OS and how much this size you need always reliable 1 2 3 gigabytes I always put in 1
09:14
2 8 but um and you've got this volume and it's showing up the Azure portal right there and it says hey it's attached to demo VM so
09:23
there's seamless interoperability between what's happening in in your Vmware client and here and resource utilization here if I wanna
09:31
take that from 944 megabytes per second to 2 gigabytes just a slider and you're done and demanded service takes care of everything under the covers there so hopefully
09:42
that was a quick demo that showed you what it is this is all GA so you can go to your Azure portal today you know go through the exact same set of steps here might take 10 minutes
09:51
instead of two and a half minutes uh but you can go ahead and go do that and we've had customers using this already and giving us a lot of feedback as this has been in preview
10:01
and again the ease of migration you have good cost savings here but the real thing that strikes us is the amount of time we're hearing customers tell up
10:11
tell us that it freeze up right this is something that for them they'd have to think really really hard like I've got to get out of Vmware like on prem how do I do this
10:19
this makes it super easy to kind of go do together with a pure and Azure and that really at the end of the day time is the most precious resource for all of us if we can free up your time
10:29
to go spend on those strategic initiatives around AI I think we're doing the right thing for for each other right so that's that's what our hope is we really would love you to try this out
10:39
and give us some feedback and we're just getting started here with this uh pure storage cloud you know everywhere you can run compute in Azure we want to make it easy to use pure Storage
10:52
block storage along with those things whether it's Vms containers etcetera so you'll see us continue to work on this and release on a fast cadence
11:01
we want you to try it out tell us more and there's a lot more coming in this partnership over the years to come thank you
