[00:00:01] Speaker A: Hey everyone, the KBR media team is heading to Las Vegas for NetApp Insight 2024 23rd to the 25 September. If you're interested in intelligent data, infrastructure, AI and unified data, plus learning about new announcements from NetApp and its partners, we recommend you register. The link will be dropped in the show notes. Hope to see you there. And if you can't make it, we've got you covered. KBI Media will be providing coverage for many of the keynotes and presentations, as well as a raft of one on one interviews we're holding with the world leading experts in attendance.
[00:00:38] Speaker B: I think at the end of the day, a lot of companies and most customers are customers first and they imbibe that. But the way I kind of think about it is, yes, one is putting customers first and serving them really, really well. And that's part of the DNA and the culture. I'm not questioning that. My point more is around are you fundamentally pivoting and building products and offering capabilities that meets their needs in the way that they actually think that they want the need met?
[00:01:08] Speaker A: This is KBCS.
[00:01:09] Speaker B: Are they completely silent?
[00:01:10] Speaker A: As a primary target for ransomware campaigns.
[00:01:13] Speaker B: Security and testing and performance risk and compliance? We can actually automatically take that data and use it.
[00:01:22] Speaker A: Joining me today is Mahesh Thiagarajan, executive vice president at Oracle Cloud Infrastructure, more commonly known as OCI. And today we're discussing Oracle's approach to the cloud. So, Mahesh, thanks for joining and welcome.
[00:01:34] Speaker B: Thank you so much. I'm so excited to be meeting you today.
[00:01:37] Speaker A: Okay, so, mahesh, I want to start with, I'm aware that you, as in yourself, were quite pivotal in 2016 around the launch of Oracle cloud infrastructure. So maybe talk to me a little bit more about this. What does pivotal look like? What's your experience?
[00:01:52] Speaker B: Look, I think I was one of the early members and joined Oracle back in 2016. We were trying to think a little bit more about what would it take to build a second generation cloud.
Before my time here at Oracle, I was at Azure. I was one of the early members there as well, building out azure infrastructure. For me, it was more about how do we think about kind of taking this to the next level? And so I was lucky enough to, one of the few people that joined Oracle in 2016, and we had some fundamental principles around saying, hey, we got to build a cloud that actually has security built in and it should be always on. We said we want to have a set of services that actually meets our customers where they are when it comes to enterprise customers. What I mean by that is even today, if you take a step back and look at how much of the cloud is actually fully adopted, and those numbers are still low for any industry that's about 1015 years in, if you see the adoption numbers less than 50%, you go, why? Usually because the products are not meeting the customer needs. And so I got the opportunity with Oracle to actually think about how do we build a cloud that serves enterprise customers, gives them the security focus on mission critical apps. For me, pivotal is a lot more about kind of leveraging the strengths of what we learned, leveraging the strengths of the company around data, and kind of driving it forward to build what we call as a second generation cloud.
[00:03:17] Speaker A: Okay, so there's a couple of things in there I want to explore. You said, build a cloud that has security built in. Now go back to 2016 for a moment. What was sort of going through your mind or the team's mind when you're sort of building this out and then sort of further to that, what does that sort of mean from your perspective or Oracle's perspective?
[00:03:34] Speaker B: First thing that we said was, look in the cloud. Everybody just was offering virtual machines at that point in time. And we said, why does the cloud only offer a virtual machine? Why can't it offer bare metal computers? And you go and think about bare metal computers, it actually goes back to how do you isolate yourself, which is the cloud provider, away from the customer, but still give them full access to a bare metal computer? And that required tremendous amount of innovation that we did around our network offload technology. The security around the network offload technology.
So that's sort of an example of what we meant by security built in. Not only that, I'll take another example to articulate every single product or service that we built. We said we're going to only offer security or capability or features that would meet our bar. So you take an object storage service, for example. We just decided we're never going to offer an object stored service that offers buckets that has to be served on HTTP. We set everything and anything on our cloud only we served on HTTPs. If you really think about a fundamental principle, we wrote down on a paper, and it just became a security tenant in everything that we build at Oracle. So those are just a couple of examples in how we do it.
Security built in is sort of about part, a couple of different things. One is it's about the processes and everything that you do around that, like meeting the bar in terms of security. And second is really around your product thinking and how you evolve and do so when you think about this notion of network offload technology for our computers, where all of the cloud and all of our services, including customer stuff, runs, when we say built in, every single piece of infrastructure that we land follows the same identical principle about this network offload technology. We don't create computers or offer anything to our customers that doesn't actually have these capabilities in it. I, and then I talked about processes, any product or feature or capability that goes through, we really spend a ton of time reviewing the designs, the products, the capabilities that we offer to our customers to meet that security standard that we set for ourselves and ensure that our customers, even if they do accidental choices on our cloud, don't end up exposing themselves to risks that they would probably be worried about. So it's more about how we think about our cloud and second, how we built security into the cloud and the bare metal cloud computers as an example.
[00:06:01] Speaker A: So what do you mean by accidental choices? What does that mean?
[00:06:05] Speaker B: A good way to articulate that is usually examples. If you think about a lot of press around security issues or mistakes or things like that, it fundamentally boils down to customers doing a lot of misconfigurations and how they use the cloud.
One of the things that we told ourselves is we got to reduce the surface area of a customer making a bad choice on our cloud. I'll take an example and you think about you getting an account on a cloud. There's a lot of regions where oracle cloud infrastructure operates. As a customer, we make it easy for them to say, you cannot accidentally replicate the data that you place, say for example, you're a european customer, accidentally replicate the data to another country that you're not supposed to, and violate some of the data sovereignty principles. Those are things that are very hard for a customer to make misconfigurations on. That's one example. The second thing that we said was, hey, if I'm a customer and you have offered me all of these capabilities and APIs and experience to develop on top of you, how are you as a cloud provider going to help make it easy for me to not make these mistakes? So we thought about a couple of products in this area. One is called Cloudguard, and then that is called as a maximum security zone. Cloudguard is this concept where you just have to add a single click of a button. It is constantly watching all the changes for your account in your tenancy, and it is telling you if you are making any accidental misconfigurations pretty fast. That is a reactive posture, but it does all of that so you dont make any mistakes. And it brings oracles hundreds of decades of experience of things that could go wrong, and it actually has hundreds of recipes that's constantly being running in your account. Second is really this concept of a zone where you just can't make a mistake. So we have this idea of a maximum security zone. When you build your applications in it, it may be a little harder, but it ensures that you don't make a mistake in it. Example, you try to create an object storage bucket and you try to, it has some important backups for your database or whatever, and you try to put it in the public. The cloud will not allow you to put that object storage bucket on a public Internet. It's just not possible. So we thought a lot about how do you build these capabilities on into our cloud? And those are two examples of products that are also available in our cloud that customers can turn on and at zero cost. Both of those features are available for free of cost.
[00:08:30] Speaker A: Okay, this is interesting, I just want to zoom out for a second. Just talk about cloud in general or cloud providers in general. What do you think people get confused about when it comes to the cloud? Because I've spoken to people and other cloud providers at similar levels to yourself, they sort of say the same sort of thing around that misconfiguration and people just think, well, I set up an account, I have to do anything after that. But I want to hear your thoughts on what you've seen with your experience.
[00:08:55] Speaker B: When you think about sort of security. Security at the end of the day is sort of a shared responsibility model, right? At the end of the day, the customer has responsibility for doing certain things. And from our perspective, the cloud has an important role to play to make sure that we do everything with all of the experience that we have to bring the best to bear to make sure that the customers are successful. So what I have observed is usually when you don't have sort of this big barrier or sort of the developers who worked in this cloud for a really long time, right, they're really trying to like hit a project deadline or something and then just kind of moving really fast to deliver on a project which is good for the business, but at the same time, you know, there's some mistakes that end up happening because it's not done out of ill will, but more about not knowing enough. So partially when I talked about Oracle cloud guard or maximum security zones or these capabilities that are inbuilt into the cloud, it's really making it easy for a developer who's building on the cloud to not make those bad choices. So what I would say is, my observation with customers is they're trying to do the right thing and get stuff done and leverage the advantage that cloud brings to them. But more so, the time, agility, speed, with the lack of experience and the difficulty that exists in other clouds make it hard for them to solve the problem and not also make mistakes. And Oracle cloud aims at backing that problem in two different ways.
[00:10:18] Speaker A: Okay, I want to focus now on, because as we've sort of mentioned throughout the interview, we're quite a pivotal person in OCI. So maybe what have you sort of learned since this journey? So 2016 was around eight years ago, so you've obviously seen a lot of change and development and evolution. So maybe talk us through that. What are the things that you've learned? And then probably my second question to that would be, what do you think then moving forward? What are some of your learnings that you'll take then moving forward into 2025 and beyond?
[00:10:49] Speaker B: I think one of the things that I personally learned was, I mean, this is gonna be sound, very philosophical, but, you know, I kind of, I think people matter, right? So my number one priority has actually always been working with a great bunch of people and kind of having a culture about, like, doing good work. So that's definitely one. But going beyond that, I think we've learned a lot about how to fundamentally hold what I consider to be a consistent bar. So to offer customers a tremendous experience in a very short period of time. Right. So one of the advantages that we have is sort of, we can, we started a little late, so we learned a lot from what other clouds did wrong. But one of the disadvantages also, you started late. But for us, taking the lessons, using that as an advantage and seeing if that actually allows us to accelerate hundreds of cloud services deploying over 70 plus regions around the world. I think we've learned a lot about how to be fast, how to actually offer all of those features, but still not compromise on quality. For me, I think I've always struggled with, hey, how do you get so much done in such a short period of time? And I think we've learned quite a few things about how do you maintain this bar while still keeping up the pace. So that's definitely one. The second thing that I say I would have probably learned more is that I think I've learned a lot more as we engaged to understand, hey, even though the clouds have been around for 1015 years, when I talk to customers, they still have complaints. So what I've learned a lot is to appreciate sort of the fundamental challenges that still exist with the clouds. And obviously that gives us an advantage in terms of like, hey, still there are unmet problems in the cloud. And so I'm always learning from our customers about how we can do a better job of meeting all of their needs as we look ahead.
[00:12:40] Speaker A: Okay, there's a couple of things which is interesting that you said in there. Now, I went to Oracle Cloud world tour in Singapore. I interviewed your CIO Jay Evans, and that was one of the questions I asked her, like Oracle or OCI started late. So you mentioned before, because you had a late start, it gave OCI the opportunity to see what the other clouds did wrong. So what would you say, generally speaking, what would you say? What was your observation about where they went wrong?
[00:13:09] Speaker B: Perhaps at the end of the day, all the clouds are successful in terms of revenue and others, but I think there's some learnings in there. First, for me, it has a lot to do with all the clouds experiment in terms of products or features or capabilities, if you will, in terms of launching new things to see how customers can adopt.
All of them have a perspective in terms of how they want to actually simplify the customer challenge. But when you do that for a while, there's nothing wrong in what they attempted. But it's, hey, would I have launched service x to solve problem y? And was that enough adoption? They probably had a, had a first start, so they had to make a bunch of attempts. But then they also know, hey, those products never took off. You know, three years later, customers are not probably happy about it or the adoption has not gone well. So the advantage that I talk about is really about, they tried on a lot of different ideas that gave them sort of like an indication of what customers really want. And that allowed us to focus on things that we did not see, we did not want. Their customers said we did not like that. So that allowed us to not experiment on things that customers don't want. So it simplified the problem area that we were going to tackle. So that's definitely one, the second advantage is we did not have what I consider to be sort of the weight of the legacy, because at any point in time there's latest and the greatest hardware innovation, latest and the greatest network offload technologies that were always available at our hands. And so for us that was a significant advantage. So I don't think they went wrong, but it is more, when you go first, you're obviously going to experiment a lot more. And you have tons of results from those experiments that we took as lessons. And we also have the advantage of latest hardware technologies and innovations. We married the two to play to our advantage.
[00:14:52] Speaker A: Okay, I'm curious to know with your experience now, obviously, you know, you've just mentioned sort of later to the game, if you had to pick a position, would you say it's more optimal to be first move? Is advantage to be like, okay, well, we'll first, or would you say in OCI's position later to the game actually is a better position to be in? If you had to choose, which one did you choose and why?
[00:15:15] Speaker B: Look, I think first movers advantage always has an advantage, right? Yes, you get a, you get, you get time obviously to play and experiment at first mover is great, but, but I also believe the market is so large from the way I think about it. So there's room for a lot of different players. And as you can see, Oracle cloud infrastructure is growing pretty well as well. And so what is obvious is at the end of the day, if the market is not fully captured first, most advantage could go either way. Right. So what is more important is how effectively are we serving the customers and how much more market demand exists and are you still best positioned to like serve that market. So for me, first mover advantage is great. I pick that. But really first, more advantage, not play to the, you know, to the success of actually capturing most of the market demand is not really first, most advantage. So, you know, you could go either way.
[00:16:08] Speaker A: So you mentioned before, effectively serve the customer. What does that look like? Because I feel like, you know, every, every company says, well, we effectively serve our customers, but what does this mean from your point of view in your position?
[00:16:20] Speaker B: Yeah, look, I think I, at the end of the day, all of, you know, a lot of companies and most customers are customers first and they imbibe that. But the way I kind of think about it is, yes, one is putting customers first and serving them really, really well. And that's part of the DNA and the culture. I'm not questioning that. My point more is around are you fundamentally pivoting and building products and offering capabilities that meets their needs in the way that they actually think that they want the need method? I'm just going to take a good example on our philosophy around everything everywhere.
We had this idea back in 2016 that, hey, we do not want ever our customers to think about how to build their applications differently in Europe versus us. We said all the capabilities that we built at launch will always be available in every single region Oracle cloud infrastructure is going to be in. So we call this everything everywhere, which means it's available. That's one example of how we take an approach of putting customers first, but we take the burden away from them. Second is our concept of dedicated regions or alloy or others where we said, look, less than 30% of the market is needs are actually met in 15 years. Clearly there is a need for the remainder of 70%. Why are they not moving? What is driving that problem? And that kind of drove us to come up with the innovation around this notion of dedicated clouds or dedicated regions that we can actually give to customers and we can deploy them in their on premises data centers right next to the servers where they're actually like literally running their existing applications. And second, we also understand that the market needs are like evolving around data sovereignty, more data privacy. And a lot of these enterprise customers face the difficulty around moving these applications over into the cloud because there is a big barrier for them re architecting. When I say customers first one, it's about DNA and making sure all their needs are met in terms of how they're using our services. The second is really pivoting your strategy, our strategy to go to places and solve problems that nobody else has attempted. So for me, that's putting customers first.
[00:18:32] Speaker A: Okay. So, mahesh, I'm really curious to know just from your point of view. Now, I have asked Jay Evans this question and you've sort of touched on it, but maybe we can get into this a bit more because I think this is really important. Now, I want to know why a big player like Oracle, who has a $385 billion market cap, wasn't as fast in the market in the cloud arena. It's only you guys don't have the money and all those types of things. I'm just, I'm really curious because we've sort of articulated, you, quite pivotal person, want to get into this because I think people in this space are asking these questions. So I think it's an opportunity now to say this was the reason why, which sort of maybe sets the scene for people, for people out there that I'm speaking to.
[00:19:10] Speaker B: If you think about like Oracle kind of making an attempt at the cloud or building the cloud, we've been in this business for a long time. Right, but I think partially, but you have to understand, Oracle is actually an entire end to end business from an enterprise perspective. Right? What I mean by that is we have a thriving data business. We are a huge SaaS business. So from serving our customers there are a lot of different businesses that Oracle is actually in the midst of and we've actually had a cloud on the bottom, but I think it's more around how do you take it to the next level? So for me, I don't know if I would categorize this as, hey, Oracle did not make an attempt. Oracle did make an attempt. But I think it's more about how do we actually continue to drive that and get the market traction at the right time. Sometimes it's how does the market see us over time, rather than, hey, did you make an attempt or not? So for me, I think Oracle's got a thriving data business, SaaS business, and SaaS actually ran on cloud, just so you know. And the SaaS business was around even before OCI started taking off. Right. So we've made attempts at cloud, but I think the market recognition came in much later.
[00:20:19] Speaker A: Yes, and you're right. And I think people that I speak to in this space all over the world, like they do complement OCI. And I think it's just perhaps the market hasn't quite understood that in any. I think that's important. Why you're here today is to really share that fidelity a little bit more about the capability. So I want to follow this a little bit more. If you could change sort of one thing about OCI, like, what would it be? Because I know you obviously sound like quite a reflective person and you've got a lot of experience in the game. So I'm curious to sort of see, like, is there anything that you sit back at night and go, oh, you know, perhaps I should have done that differently.
[00:20:54] Speaker B: Every leader at any point in their time, if they actually tell you that they don't want, they probably would have done things differently. They're probably lying. What I can tell you is this. My fundamental philosophy on just general problem approach is like, you are always collecting as much data as possible to make the best decision possible at that point in time. And usually six months later you could go, well, I have this other piece of data. I should have done this right, but you don't have hindsight. 2020 generally, you're making the best decision. The second more importantly for me is don't make a decision when you don't have to. So those are some fundamental principles. Now, in terms of OCI, my personal perspective is that more than thinking about sort of changing a few things, I think we've got a sort of perspective in how we treat customers and how we think about tackling the market and their needs a little differently and how we execute against it. What I would probably be more focused on, and it's actually what I focus on every day, is about how we don't lose the superpowers that we built over time, that as we go and grow as a massive business. Right. So that's been my big focus. So the thing that I wouldn't change is an easier question for me to answer on changing.
[00:22:02] Speaker A: One thing you said before, don't make a decision when you don't have to. What do you mean by that?
[00:22:08] Speaker B: There are decisions that we make generally which are one way door decisions. What I mean by one way door decisions is you're probably driving a lot of money or capital into making a project investment or a big bet for launching a capability or whatever. And when you make that at a time when you don't actually have to drive that capital, but you can actually make inch towards driving that capital instead of going all in, that for me, is sort of thinking about that as a one way door decision versus a two way door decision. A two way door decision is like, look, we can take a call, but we have sufficient room and time to pivot or change, or change our strategy without actually paying a lot of opportunity costs. So a one way door decision is something that I think of as sort of like, you don't have to drive a decision. You think you spend a ton of time doing critical thinking before you enter that door. That's a different way to articulate the same problem.
[00:22:58] Speaker A: Yeah, that's interesting. Would you say that people just make a decision and sort of see your point around the door analogy? Like, perhaps they just do the one way door, like they're not sort of thinking sort of on all angles. What were your thoughts then on that?
[00:23:13] Speaker B: Not necessarily. I think for me it's more about kind of spending the time thinking about what kind of decision I'm making, if you want.
So trying to say, hey, look, there are a few decisions that I take on a daily basis or a weekly basis or whatever. Just kind of recognizing what decision type you're taking is important. That's where if you spend that time, because not always are you taking a lot of one way decisions. Sometimes it's very few. So it's just recognizing those few pivotal moments when you're doing that decision, just recognizing it and saying, hey, how do I get more data? Or how do I think more about more? How do I put the business or our customers first in making the right choice for them? Right? And that is what I mean. But usually most decisions are places where you can pivot and learn. So one way door decisions don't come often, but it is something that is important to recognize as you're making a decision.
[00:24:04] Speaker A: So I want to sort of understand a little bit more. As I mentioned, I've spoken to a number of different cloud players out there. What do you think rattles people about the cloud? Now, I asked that question because I'm speaking to certain people, like whether they're outside of it, et cetera. They seem to have all these questions about the cloud, and it seems to really get to them. So from your point of view in speaking to customers day to day, is there anything that sort of stands out when I ask that question?
[00:24:30] Speaker B: If you roll back the clock, what probably is 1012 years ago, if you will, I think there was a lot of questions about, hey, what is the cloud? Why should I move to the cloud? Cloud is actually going to, what value does it add for what I do today?
I think there were a lot of questions, I would say even about seven, eight years ago. But I think as we move ahead, it's obvious and clear that the value that cloud brings in terms of bringing the latest and the greatest technology, giving you the pay for use, actually, to some extent, some things that people don't recognize is that the security levels of what you do in the cloud are generally higher.
The challenges that customers still run into, though, is this concept around control.
And when I talk about data sovereignty, it's just an example of control.
How do we make sure that if I'm a government and I'm running on the cloud, how do I make sure that all the citizens data that I actually have, or extremely business critical information, or HPI, how do I keep that in a place where it's actually inside the same country that I'm supposed to, to meet our privacy laws? So what I would say is, I don't think cloud is a question anymore. But the forms of cloud and control is probably what people are asking more about as they understand and look, the geopolitical situation, data sovereignty, is continuing to evolve.
I think this is where Oracle spends a ton of time thinking about, hey, we understand data. We understand what customers mean when they actually run our data products on their, on premises. So it gives us an advantage to peek into their eyes and say, hey, what are you worried about?
That is what I hear more. It's a lot more about control and less, more about the cloud.
[00:26:13] Speaker A: So what do you think is the next version of the cloud? What do you see that sort of coming up on the horizon, yes, with OSR, but just cloud in general, where do you think was sort of heading.
[00:26:21] Speaker B: As an industry over time?
I do expect more and more things, more and more decisions or like software or hardware, whatever it is that enables what I consider to be like mankind or apps are going to be happening closer to where people are. Right. What that means is, hey, expecting to have, say, 70 or 100 regions from Oracle cloud or similar numbers from other clouds, it's not a lot of numbers when you think about the number of cities and people in the globe. And so I do expect more of the edge and more things happening in sort of like a decentralized way, which actually has multiple smaller installations of the cloud. I do believe that is going to be a reality both from a sovereignty perspective or just purely from like, hey, how we expect our end customers to interact with the servers and the applications that we're doing. So I do expect the cloud's going to be more, in more places and more cities than I think what we know today.
[00:27:19] Speaker A: So I want to sort of switch gears now and talk a little bit more about your team. So I'm aware that you and your team are devoted to constructing a distributed cloud infrastructure that offers customers a comprehensive, high performance cloud computing environment. So maybe talk us through this and what this looks like.
[00:27:39] Speaker B: It fundamentally boils down to what I consider to be our region deployment models and essentially tailor making, still offering all the capabilities but meeting the stringent requirements of all of these different types of customers.
What you have is we have a hyperscale public cloud, and we have a ton of customers growing using that. It's definitely something that we compete with all of the other major cloud providers in parallel. But then you go to, hey, we have government clouds that are very focused on serving different government needs, be it for various, different certifications, various different security processes. So we obviously have that. Then we have this idea that, hey, we have countries that are sovereign, or EU for that matter. So we have an EU sovereign cloud that actually has operational model that's completely operated by EU sovereign employees. So that's a different model, dedicated regions where you have customers like Vodafone or NRI in Japan who are essentially saying, hey, look, I would like to have my own set of regions, whereas the only customer is me. And here's sort of like how I want them distributed globally. And we have large customers who are making those choices. Then you have partners, right? If you look at many different partners, I mean, NRI is one of them. We have other partners in various parts, like we have partner in Italy. All of these are partners who are saying, look, in my country, I want to essentially be a primary technology service provider where we can actually take Oracle cloud, but we want to be the person that is actually managing the cloud and selling the cloud to our end customers and actually making them move their apps onto the cloud. We really think about how OCI is built with the right fundamental principles and a really small footprint to extremely large footprints. We're able to cater to these markets based on what our customers want, be it an enterprise customer, be it a government customer, be it sovereign customers, be it partners who are wanting to be cloud providers. OCI is enabling all of that while still being a major hyperscale cloud provider. And then that's the advantage when you think about this distributed cloud strategy that we got and what we're doing.
[00:29:48] Speaker A: So what do you think happens from here? What's sort of your plan? What's your strategy? I know we're sort of halfway through 2024. What can people sort of expect from OCI now? Moving forward?
[00:29:59] Speaker B: I'll talk about things that are public. What customers can actually expect is we're continuing to grow globally with more and more customers, and they should expect our OCI being available in a lot more places and definitely expect a ton of innovation that they've seen from us over the last eight years.
[00:30:20] Speaker A: So when you're saying growing globally, which I understand, what's sort of your approach to cloud adoption for OCI over sort of competitors, what's sort of your plan for that?
[00:30:32] Speaker B: If you take a step back and think about the example that I talked about, Oracle Alloy, which is where partners operate our cloud, and they actually end up serving in customers as more and more partners come in. With all of these different countries across the globe, that gives us an opportunity to expand. So that's definitely one and probably heard about our multi cloud sort of announcements and partnerships that we have with Microsoft and Google right now. And so we expect to serve more of those customers with all of our significantly differentiated data products where mission critical customers depend on Oracle. Those products are starting to be available at Oracle database at Azure, or Oracle database at Google Cloud. So you're going to see a lot of that. I think there's a tremendous market demand for AI infrastructure and we're continuing to differentiate quite a bit on that so people, customers can see a ton of innovation coming over in that space. And lastly, everybody talks about generative AI and there's a lot of exciting, cool technology work that's happening there. And for us, we're embedding generative AI on all the proper use cases. That adds value throughout our stack, be it our internal processes, be it on the cloud, making it available to our external customers, integrating that into our oracle health products, making it available as part of our SaaS. So a ton of innovation that we've done and worked on is now also being infused across our entire stack and product. So globally, we're going to grow our oracle allied partners, we're going to grow with them. We're doing multi cloud partnerships. AI infrastructure market is growing. So a lot of that work is what our adoption globally looks like.
[00:32:11] Speaker A: Just talk quickly a little bit more about AI infrastructure. What does that mean and what does that look like as well?
[00:32:18] Speaker B: Look, what I think of as AI infrastructure is fundamentally the GPU infrastructure that we actually build and join with our partners like Nvidia.
Also a ton of innovation around the networking technologies that allows extremely large language models that are the innovation that's happening in the industry, be trained on the cloud. And so we're doing ton of work on that. Not only that, we're also doing quite a bit of work in making it easy for our dedicated region clouded customer, or all of those different customers who have dedicated regions to also get access to GPU's in their own environments, small and large, and actually also leverage our generative AI technologies to train their own custom models. So when I talk about AI infrastructure, it's making those technologies available, fundamental innovation on networking and lastly providing them with the high performance storage that is required to run some of these AI workloads. So that's a lot of the work that we do. And when I talk about infrastructure, that's.
[00:33:24] Speaker A: What I mean, high performance storage. Talk to me a little bit more about what that means then as well.
[00:33:30] Speaker B: When you think about storage needs for an AI workload. So in traditional terms, people have leveraged file storage to actually, if you think about training a large language model based on all of Internet state the web data, or you look at some of the next generation models working on video or text or pictures, those all require tremendous amount of storage. And all of these large GPU clusters that are trying to train themselves on these models require high frequency access to the storage as they're calculating these mathematic equations and storing all of the intermediate results. You need a storage platform that actually serves this at the performance scale for what are these tremendously large GPU clusters offering? A storage platform that solves this need is what I mean by high performance storage.
[00:34:17] Speaker A: I asked that question because I feel like storage is one of those things in my experience of interviewing people. It just feels like it just gets relegated a little bit, like people forget about it. Perhaps it's not as interesting as other things.
Would you agree with that sort of statement as well?
[00:34:31] Speaker B: I think all of the storage experts and industry experts and the people in my organization will feel sad to hear about that. But I would say it's more of storage is sort of a given in many people's minds that, hey, the innovation is actually done, or this is what you can expect. But the way I think about it is there's tons of innovation happening in the industry around storage and next generation drives. There is tons of innovation happening around sort of tiering and laying off storage for AI workloads. So there's lots of innovation that's happening. But I think it's sort of a given at the point that people expect storage to work, so they don't talk a lot about it.
But I don't think that is true because I think the amount of innovation that we do, like, for example, one of the ways in which we built our block storage platform, we're the only cloud that offers performance SLA's, because we're very proud of how we thought about it, how we built it. We not only give customers availability SLA's, but we also guarantee them an IOP's per gigabyte. Sort of SLA. Right. So there's lots of innovation that's happening, but it's sort of not like front and center like AI, but I think there's cool work happening, in my opinion.
[00:35:44] Speaker A: So, mahesh, do you have any sort of closing comments or final thoughts you'd like to leave our audience with today?
[00:35:49] Speaker B: Well, first, thank you for the opportunity to talk to you. I'm super excited and I'm glad we made it here. I think there are four hyperscale providers and Oracle is one of them. And my ask of customers is definitely look at Oracle Cloud as you're thinking about assessing your needs, and there's always cutting edge innovation and work that is happening in our space. And I would just say, hey, you know, just take a look. Spin up a free to your account. It's always free. You can play with Oracle Cloud without spending a dime. And there's ways to get started very easy.
This is KBcast, the voice of cyber.
[00:36:32] Speaker A: Thanks for tuning in. For more industry leading news and thought provoking articles, visit KBI Media to get access today.
This episode is brought to you by Mercsec, your smarter route to security talent.
[00:36:46] Speaker B: Mercsec's executive search has helped enterprise organizations find the right people from around the world since 2012. Their on demand talent acquisition team helps.
[00:36:56] Speaker A: Startups and mid sized businesses scale faster and more efficiently. Find out
[email protected] today.
