VMware Explore 2025 | Purnima Padmanabhan, Tanzu Division, Broadcom

Clips
More from VMware Explore 2025

Purnima Padmanabhan

General Manager, Tanzu Division

Broadcom

play_circle_outline VMware Explore 2025 event introduction featuring John Furrier and Paul Nashawaty.

play_circle_outline Tanzu's Transformation: Innovations and Growth as an Independent Unit Under Broadcom Amid Strong Customer Support

play_circle_outline Emphasis on radical simplicity and focus on application and data layers.

play_circle_outline Improved security and automated remediation features for mission-critical environments.

Info
Transcript

Purnima Padmanabhan, Tanzu Division, Broadcom

Purnima Padmanabhan

General Manager, Tanzu Division Broadcom

In this VMware Explore 2025 interview, Purnima Padmanabhan, general manager of the Tanzu division at Broadcom, joins theCUBE’s John Furrier and Paul Nashawaty to unpack how Tanzu is accelerating application and data innovation in the enterprise. Padmanabhan sets the record straight on Tanzu’s role as an independent business unit inside Broadcom, detailing how its sharpened focus is fueling customer traction and partner growth. She emphasizes the power of “radical simplicity” by separating core infrastructure from app and data services, and explains why this c... Read more

explore Keep Exploring

What is the current status of Tanzu and what new technologies are being introduced? add

What is the current status and organizational structure of Tanzu as a business unit of Broadcom? add

What is the current status and strategic focus of the business as discussed in recent advisory board feedback? add

What recent enhancements have been made to the platform regarding security, vulnerability analysis, and support for different types of applications? add

bolt Powered by CUBE AI

Purnima Padmanabhan, Tanzu Division, Broadcom

search

>> Welcome back everyone to theCUBE live in Las Vegas for VMware Explore 2025. I'm John Furrier, host of theCUBE with Paul Nashawaty, lead analyst at theCUBE Research and AppDev data platforms, and of course on all the action around the developer community, of course here at VMware. VCF-Nine Tanzu bringing new technology out and great product releases here. And we have a great guest here, Purnima Padmanabhan, general manager of Tanzu division back on theCUBE to clarify and tell us about the new technologies. Great to see you again.

Purnima Padmanabhan

>> Great to see you. - Welcome back.

>> Great to see both of you. - Purnima, great to have you on

Purnima Padmanabhan

>> last.

>> You've been a legend on theCUBE. You've very articulated, you know the tech. A lot of people were skeptical about what was going to happen this year. VCF-Nine, they delivered good product results. The business results were good. Tanzu is continuing to do its thing and a lot of people want to know what's happening. So let's get into the state of Tanzu first because I want to clarify, I want you to clarify what's happening. There's been some things with Kubernetes and some movement on your side, of course the great news. So start with the status of Tanzu. What's going on? Get the scoop.

Purnima Padmanabhan

>> Okay. Let's start with fundamentals. Tanzu is a business unit of Broadcom, independent business unit of Broadcom. As a general manager, I run that business unit and I report it to Hock. Tanzu is doing very well as a business. We have excellent customer traction and we have really a strong rationalized portfolio set. And maybe what I'll also address it down is over the last year we have refined and focused as is the key with anything that Broadcom and Hock does, the Tanzu portfolio set. So Tanzu always had three pieces. There was Tanzu CloudHealth, which we now still own, but we sell through Arrow. Then there was Tanzu Data, which we have really beefed up with Tanzu Data Intelligence. I'll talk more about that. And then there was Tanzu Platform. Now, Tanzu Platform had two components in it. There was a component that has been very strong and thousands, millions of workloads run on that platform is our Pivotal-based Tanzu Cloud Foundry, heritage part of the platform. And then we had Tanzu that had Kubernetes in it, and Kubernetes runtime and Kubernetes management. Now what we did is we fundamentally Hock said, "Look, Kubernetes and Kubernetes management is core infrastructure and it needs to be together with the infrastructure, while the AppDev pieces should stay separate and independent in Tanzu." So what we did early this year is took all the Kubernetes and Kubernetes runtime pieces and put it into VCF. So when you go into VCF-Nine, doesn't matter, containers or VM, you get one common platform. When you get to Tanzu, you get apps and data platform. Simple. Runs on any IaaS, it is clean and it is the fastest way to get your applications to production.

>> And I want to point out, because a lot of people get confused because they see the feature, all the VCF focus because that's the primary mission on the enterprise cloud. But Tanzu isn't a sideshow, it's part of Broadcom and it's intentional. This is an intentional moves that are being made and your business is thriving.

Purnima Padmanabhan

>> Our business is thriving. And in fact, I can't tell you, it's almost until you go through it, you don't realize it. There is an amazing power to simplicity, radical simplicity and focus. So we are not focused on any infrastructure components. We are 100% focused on application and data layer. We just did our advisory board with customers and they said, we are blown away by the level of innovation you have been able to put in. If you were stretched thin, we could have never put that type of innovation. And so then I also get this question, which is, hey, Tanzu wasn't featured much in VMware Explore main stage and so on, which is kind of what you were referring to earlier. Well, Explore is fundamentally about making the private cloud happen. Once the private cloud is there everywhere you can build additional layers on top of it. And so it's again about focus of message. What we want to get out as Broadcom to the customers is private cloud is the best, safest, cheapest, easiest way to run your .

>> It floats all the boats, and we-

Purnima Padmanabhan

>> Floats all the boats. Exactly.

>> We had Krishan who runs VCF say

>> with a customer, we're just a plumbing. We're enabling the good stuff to happen. Where AppDev comes in, this is where I think it's important to kind of squint through and understand the nuance. You win either way.

Purnima Padmanabhan

>> When VCF wins, we win because ultimately, once you have a great private cloud, you want to create this fat pipe of applications and data pouring into the cloud and that is where Tanzu comes to.

>> Yeah. Absolutely.

Paul Nashawaty

>> And I think I want to set the record straight. Last year we were at Explore, we heard there was what's going on with Tanzu? Is it going to be divested? Is it going away? Is there no innovation? It's so far from the truth now. I mean just coming on the show for-

>> Two and a half years in, we are strong.

Paul Nashawaty

>> We're strong, and yeah, it's bifurcated and it's split.

Purnima Padmanabhan

>> And it's shown that the Kubernetes piece going into VCF like you're talking about that really made that foundational platform. But that developer focus and the data intelligence piece is really where you're doubling down on. This is something that most people are not aware of, but it was something that's core to the product initially. Right?

Purnima Padmanabhan

>> Absolutely. So if you think about Tanzu's heritage, the core asset came from Pivotal. And what we did very well at Pivotal was we knew the developer, we knew the apps and we knew the data teams. So I'm talking about millions of apps running on Tanzu Platform. I'm talking about petabyte scale data running on our intelligence solutions. So these are not just new, these are fundamental and at the heart and core of many enterprise businesses. And what this allowed us to do is in a way we had contaminated the Tanzu brand with infrastructure as well as app and data. This allows us to say that it doesn't matter where, let us get a strong private cloud. And then the best way to put apps and data on the private cloud becomes Tanzu.

Paul Nashawaty

>> Let's talk about execution though, because I think execution, the heritage is solid. Everyone now knows this is not going anywhere, you're investing, you're innovating, now looking at the future. And when we start looking at MCP and we start looking at agents and we start looking at the connection of the intelligence of the data to the applications, that's what's new. That's what's really coming to life here. And I mean when I walk in the show floor, when I'm sitting in the sessions and I'm talking in your session, it was all, you could feel the excitement in the room about this is where it's going.

>> Let's get into the announcement. You have news,

>> you've unveiled new things here at the show. So it's not like you're, of course you're doing things. What is the core news that was revealed?

Purnima Padmanabhan

>> So let me connect both the dots here. I think as I go into the news, let me define why the news. Okay?

>> Yeah. - AI has truly created a jump ball for application

Purnima Padmanabhan

>> and data stacks as never before. AppDev has gone from IDs to coding assistance. App stack has data from distributed pools of data to data lake houses with clean interfaces and analytics from individual stacks to foundation model- based rapid analytics, right? So everything has transformed and we believe the winners in the enterprise are the ones that bring app data and AI stack together with safety, governance and compliance. That is what Tanzu does. So our mission has stayed the same. It is all about accelerating business velocity by accelerating application velocity. And we do that by helping ideas to code and code to production. Now, there are two pieces of news that help that. The first one is we introduced our Tanzu Platform, 10.3. So the Tanzu Platform allows you to very rapidly build applications. We not only help you develop the applications faster with clean guardrails, we give you the whole paraphernalia for a microservices-based applications, the application stack, the connection to the models, the connection to the tools and MCP servers, the connection to the vector databases all in one place, fully made sure that you can bind it so that a developer can be out to production in hours, not in months and years. And that is what is Tanzu Platform. And with 10. 3, we have done a few things better. Number one, we have made it very even easier for the... We have bought AI to enterprise app developers. I don't know, John, if you remember, even in the past we have talked about it with Spring AI and with Java in Spring AI. And we have been able to bring more of the developers to the AI fold rather than AI sitting in an ivory tower somewhere. So first thing is a lot of capabilities, whether it is model brokering, and this model could be running on private AI foundation in VCF by the way. So that is an important theme. Or it could be running in public cloud. So whether it is model brokering, whether it is connecting to your models and tools, whether it's RBAC, rate control, governance, all that is built in and the developers simply connect and they're off to races. The second thing that we have done is we have really expanded the security and vulnerability analysis and automated remediation on the platform because what stands out separate from public cloud is as people think about sovereignty, as think people think of controls, we run in the most critical environments, mission-critical environments like banks and healthcare and federal systems. Having that top to down easy panel for security is very, very important. The third thing, and I'll tell you only three things here, the third thing that we did is we always are fantastic for greenfield apps, but now with the latest Tanzu Platform, you can point to any brownfield app. We simply onboard it, rationalize it, and make it modernize so that you can put AI into it. So that is Tanzu 10.3, the first part of the announcement. Question on that, then I'll go to the second part.

>> Is there a lake house in the platform? Is that part-

Purnima Padmanabhan

>> So that is, you're coming to the second part of that now.

>> Okay. All right. So platform, that's an enabling platform.

>> So Tanzu Platform is an application platform.

Purnima Padmanabhan

>> It helps you rapidly build applications and run them at a production scale. But then with AI, everybody's been staring, okay, I can get a model, I can get a database, I can write code, but this intellectual property that I have in data is a mess. It is sitting in SQL databases, Oracle databases, file systems, S3 storage, and I have no clue how to make sense out of it. And that is where the second announcement comes in. So Tanzu Data Intelligence gives you a data lake house that can be put on-prem. It is based on our very strong, massively parallel data warehouse called Greenplum, which is already in production. And we have extended it to be a lake house by including file systems so that you can now hold structured data as well as unstructured data. It has got full federation capability, so it can even federate out to S3 or federate out to your old HDFS file system. And then we put a single query layer on it, and imagine you can have a SQL query to talk to this unstructured and structured data. And not only that, a data warehouse is good, but you need to have a way to rapidly ingest and put information into it. So all the data flow, the transformations that you do is also included with that in data intelligence platform.

>> So the lake house brings data in from

>> the disparate sources?

Purnima Padmanabhan

>> Disparate sources and federates into them and gives that common query. And the most important part is once I have a lake house, it should not be sitting in a corner just for some reporting. I need to use it in my AI apps, I need to use it in my actual real-time apps. So we have a streaming output with caching servers like GemFire, Redis, and Postgres that we can automatically put on the cache for you so that you can use it with your application.

>> And one thing to kind of clarify,

Paul Nashawaty

>> so we're clear on this announcement, it's an important announcement to make sure that people understand, it's you're not going after people like Databricks or Snowflake. You're actually working with these people so you can actually have the data on-prem to access that. Is that correct?

Purnima Padmanabhan

>> I think the thing is from a paradigm perspective, Snowflake and Databricks are trying to do the same thing, but it's a very different paradigm. They're trying to put everything in the cloud. They're saying bring all the data to me, put it in the cloud, and then query and so on. What our value proposition always has been, look, VMware, Broadcom brand has been private cloud brand. And what is amazing about this data intelligence is imagine I could give you the power of querying, the power of high velocity as well as archival data, the power of structured and unstructured data, the power from going from few terabytes to petabytes of data and being able to use it in real-life applications on-prem. And what it is, it is sovereignty, it is safety, it is security, but the biggest one that comes up, cost and cost predictability.

>> Of course, of course.

Paul Nashawaty

>> I also think that if you combine the two pieces together, the explosion of application development is occurring. You see this, having the Tanzu pieces, application creation, and you don't necessarily have to have a professional developer doing the development. You can actually have the lines of businesses working in this space. Not to say it's a low-code, no-code. So that's not what I'm saying, but it's more enabling the lines of business to actually develop and access that information.

Purnima Padmanabhan

>> That is exactly right. That's the vision. So not only have we announced Tanzu Data Intelligence and Tanzu Platform 10. 3, these are integrated in turn. And by the way, these two also run very well on VCF. So Tanzu Data Intelligence is optimized for vSAN Max, VCF- Nine, and then Tanzu Platform runs on VCF-Nine. But these two are integrated. So I'll give you a very nice example. A customer was saying, I'm trying to build an AI app. Okay. Great. You ingest the data first because you need data for AI.

>> You need data. Yeah. - Put it into the data intelligence

Purnima Padmanabhan

>> platform, then you want

>> to do transformation and vectorization. So you go to Tanzu Platform, use your models there, you vectorize the data, and the model could be on private AI foundation, but you are vectorizing the data and then you put it back into the lake house. Now if I want to do inferencing, I don't hit the model again. I simply do a similarity search against my data warehouse and do inferencing at a very high, high speed. So these are the examples that are beautifully coming together for us. As I said, AI has been an amazing unifier, but as we bring AI more into the real enterprise apps, people have realized I don't need a separate AI platform sitting on the corner. I need it to be mainstream.

>> Is the problem you're solving, connecting the data discipline to the app side or app to the data discipline?

Purnima Padmanabhan

>> It depends on where you start. It's both.

>> Well, because I think this, I mean a lot of people are doing rag apps, augmented generation, and those are the people just coding up. So they need the vector embeds, they're going to need all that search, they're going to need data. But you're saying that you guys can solve that problem if it's disparate, I don't have to wrangle it. I don't need to build-

>> You don't have to.

Purnima Padmanabhan

>> Exactly. And so look, AI apps are... The need to put AI into enterprise apps is what is driving this discussion, right?

>> Mm-hmm. - And then each business unit is trying to say,

Purnima Padmanabhan

>> "Oh, I want some data set. " And then they make a copy of that data set and now it is obsolete and so on. So what we are saying is imagine we do both. We give you a beautiful platform so that you can build your AI apps, iterate through models. We even have a plugin into coding assistance that help you do guardrails while you're building code and build write code. So you're doing all of that. But then as a business unit, if I want a data set, I just go to data intelligence, I ask my data team to give me a data set and they provide it cleanly.

>> The benefit is

>> that you just make the apps better for the customer. All right. So I buy that. So I have to ask you a question. Okay. If I'm a customer or I'm someone that says, okay, I want a platform, the trend in the industry is re-platforming things, the security companies are doing it. Having a platform that's easy to use, has all the data, is definitely a good thing. So love the approach. Okay. Now inside the company, who do you talk to that makes it happen? Who operates it? You mentioned the data side and then the AppDev side. What has to happen? Can you take me through the use case? Okay, I'm in, you sold me, let's go. What do we do? So do I call my AppDev guy and then my data person? How does it work? Who do you call on?

>> Traditionally, it all depends on the type of enterprise.

Purnima Padmanabhan

>> Traditionally what we have done is we have told to sold to two personas. One persona is a platform persona who buys the Tanzu Platform. You have known this for a long time, and the platform teams that curate experiences for developers and that persona is still absolutely the persona. But we have also been selling to the data architects and the data warehousing specialists. But what has happened is with AI, and many times they've always converged at some point, right? We are going one level up in the conversation. In fact, today I was talking to a customer in the telco space and they are trying to wrangle their data to solve for an app problem. And suddenly they're saying, "Hey, where's my data team? Come to the table. Where's my platform team? Come to the table." So our conversation is going one level higher where these two meet. And the point is, how can I do it in a systematic, consistent way?

>> And everyone I talk to that's doing data transformations, they realize that it's disparate data. So the data teams all know this, but now they need to come to the table and do something. The benefit is once it's done, that's where this platform shines.

Purnima Padmanabhan

>> Correct. And the big thing, big value proposition for the data architects is rather than everybody creating their own data set and putting under their tables everywhere, I in fact have coined a new thing, which is basically it's the data set as a service. So it's not database as a service, who cares? That's just a place where you store data, it is the data set. And imagine each business unit needs a particular data set to work with, or each app needs a particular. So imagine we can serve that data set as a service. Now the data architects are actually much more in... It's back to that control plus flexibility problem that was there in that case.

>> So the outcome then, I guess, Paul,

>> you've got data on this, I'd say would be more apps are being built.

Purnima Padmanabhan

>> More apps are being built. - So that's a

>> key metric, right?

Paul Nashawaty

>> Yeah. - More apps are being built. They're

>> good apps, the data's there. I'll say the data is the apps and it's just running. The data's all automated from that point once it's connected.

Purnima Padmanabhan

>> Then the apps and data are run automatically for you.

>> And to your point though, I want to bring in, okay, most of the people we talked to are like, okay, I got Snowflake for the analytics, I'm using Databricks for some stuff. I got vector search and this other group, there's a lot of other lake houses, there's other vector stuff. So how do you rationalize that with a customer? What do they have to do? They have to stop doing this or how do they integrate the Tanzu Platform? This is just another federated link or-

Purnima Padmanabhan

>> Yeah. So Tanzu Platform is an app platform. They can do that. On the Tanzu Data Intelligence side, what we say is you don't have to boil the ocean. Often if there is a use case, start with the use case. But the powerful piece about Tanzu Data Intelligence is it doesn't have to all sit in my lake house. We have a very amazing capability to federate into other data sources and make meaning out of it. And allow the use cases to grow and then the data to actually move in as it gets transformed-

>> Mechanisms for the developers to, if data's already kind

>> of been set up, you can just manage it's different by account.

>> We can manage the caching of the data and each team can

Purnima Padmanabhan

>> get a different data set that is cached for them and they can use that.

>> You're versatile. - Yeah.

Paul Nashawaty

>> And it's also important to note you reduce the complexity.

>> That is a key factor here. When you're looking at these new modern sets, it's not about a bag of bits as used to be talked about. It's really about here's the solution and you don't have to cobble it all together yourself. You already provide that.

>> Exactly. And the other, you asked an important question,

Purnima Padmanabhan

>> how do you get started, right? I tell my customers, you don't have to boil the ocean. You need one business unit, one app that needs to get to production. And we will show how do you get the data for the app? How do you consolidate it? How do you make it available? How do you instantiate the model? How do you get the app and get it all going on VCF-Nine?

>> All right. Well, great to have you on. Give a little indicator on how business is going. What are you optimizing for today? What's your plans? What's your plans for the rest of the year? And next year here, what do you hope to say on theCUBE in terms of accomplishments and launches? I mean, you can't really reveal your product roadmap but-

Purnima Padmanabhan

>> Exactly. - ...

>> you can give a taste, give us a feel for the vision.

Purnima Padmanabhan

>> I think the business has really, with the focus, I can see a huge energy in the team, in the partners. And especially I think as VMware, we have done some disservice to our core platform customers, to our core data customers, the Greenplum customers, et cetera. And they are really enjoying that focus in the core enterprise space. So from a business perspective, we are very doing very well. Key component, great financial results, which I can't of course share because we don't do individual BU-level results. And we are also-

>> But notable.

>> Notable, absolutely. Notable

Purnima Padmanabhan

>> and very much on a good positive trajectory. And the other piece is that we are focused, just like we focused our platforms and data, we have also focused our accounts that we go after direct and the rest of it is with partners. The other big thing that I've done is rather than doing my own services, I used to do a lot of Tanzu lab services. I am actually depending a lot on partners for services. So maybe here's a call to all the partners to join this fun innovation process that we are in. Help us in the customers. If you are in the app space, join us with that.

>> It's a growing ecosystem. Purnima, thank you so much

>> for clarifying and sharing the news. A lot more to talk about agent models, model servings. A lot more going on in the AppDev world is continuing to change. There's so much growth opportunities ahead. As developers get their hands on the data and as new ideas and new opportunities with process and workflows come to the table, just a great opportunity to enterprise. I'm John Furrier with Paul Nashawaty. Thanks for watching. From the-