On this episode of Six Five On The Road at Pure//Accelerate 2024, we are joined by Pure Storage‘s Prakash Darji and Shawn Hansen, GM Digital Experience Business Unit and GM Core Platforms, respectively, for a conversation on Pure Storage’s latest digital experience initiatives, storage solutions, and AI strategies.
Our discussion covers:
- Insights into Pure Storage’s latest digital experience initiatives
- The impact of evolving storage technologies on enterprise IT strategies
- Pure Storage’s approach to sustainability in data management
- Challenges and solutions in current data storage trends
- The future of data storage and Pure Storage’s roadmap
Learn more at Pure Storage.
Watch the video below, and be sure to subscribe to our YouTube channel, so you never miss an episode.
Or listen to the audio here:
Disclaimer: The Six Five Webcast is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors and we ask that you do not treat us as such.
Transcript
Patrick Moorhead: The Six Five is on the road at Pure Accelerate 2024 here in Las Vegas, my second home. I love it here. Daniel, how you doing, my friend?
Daniel Newman: It’s great to be here. Pure Accelerate 2024, Pat. It’s a big moment in inflection. We’re seeing all of the infrastructure discussion starting to come to life as companies are trying to find the value in their enterprise investments. I mean, look, a lot of CapEx, but now it’s consumption. Where’s this AI being consumed? And of course, data platform, storage, cyber resiliency. These are big topics and big topics here at Pure Accelerate.
Patrick Moorhead: Yeah, I mean, industry inflections we’ve seen where it’s, “Hey, infrastructure is cool”, and then, “Oh, it’s software. That’s cool.”
Daniel Newman: It’s all cool.
Patrick Moorhead: But the great thing about AI is infrastructure is back in focus again. And by the way, companies that have infrastructure platforms leveraging hardware and software seem to be differentiating themselves, whether it’s in the GBU space or even in the storage space. And to continue this discussion, Prakash and Shawn, welcome to The Six Five.
Shawn Hansen: Thanks so much.
Patrick Moorhead: First time on The Six Five. Congratulations to both of you. You had some big announcements here and your teams are making it happen.
Shawn Hansen: Thank you. Great to be here.
Patrick Moorhead: Yeah, you got it.
Daniel Newman: Yeah, it’s good to have you both here. You heard our preamble. I mean, look, we’re kind of finding that inflection this year. Last year, all this enthusiasm and excitement, you know, LLMs, GenAI. But the truth is, for the enterprises, it’s really about finding value. It’s about being more efficient, more productive, getting access to your data. And of course, for the vendors, it’s about making this really digestible and consumable and available.
And that seems to be what you all did here. So Shawn, I’m going to start off with you and talking about Pure Fusion. This is a big moment. This is one of the items I think can solve a lot of the challenges for companies trying to get more out of their data. Talk a little bit about Pure Fusion, the announcement, and why it’s going to solve some big problems.
Shawn Hansen: Great. When we first introduced Pure Fusion, we were so excited about creating this unified global pool of storage. But we realized for about a year in market that we had several things that kind of held it back. Number one was customers wanted it to work for all of their existing data.
Daniel Newman: Right.
Shawn Hansen: It wasn’t enough just for new pools of storage or new things being deployed. Customers asked, “What about everything I have right now?” And so the big announcement we made today is that Pure Storage with simple a software update will work backwards compatible with everything that you have. So really, it’s an announcement about being primetime, that it’s available as a simple non-disruptive update for all of your environments, and now you can create a single control plane for all of your appliances.
Patrick Moorhead: Yeah, we had a great conversation with Charlie, really hammered on the accessibility part, right? Accessibility to all; of course accessibility for AI and folks like that.
So another huge trend here … And it’s funny. We always think that cybersecurity is something new, but it’s not. I mean, you look back 40 years …
But as we’ve disaggregated infrastructure, we’ve disaggregated even applications, the threat matrix goes up. And Prakash, what are you hearing from your customers related to cyber resiliency? I mean, gone are the days of perimeter, “Hey, we’re just going to keep them out. Nobody’s going to get in. Therefore, we don’t even have to have a plan to get them out and protect our data.”
Daniel Newman: There’s someone reading your email right now.
Patrick Moorhead: Exactly. Hopefully somebody can respond to it intelligently. No, but seriously, what are your customers telling you about cyber resiliency?
Prakash Darji: Well, it’s changed a lot in the last 12 months. I think we’re on the early days where people are still understanding that perimeter security isn’t enough. There’s … I don’t know. It’s a fragmented market with security tools, and everyone’s like, “How do you protect the network? How do you isolate the database? How do you do security and rolls over here?” But at some point, you need to make sure every layer is protected, and storage needs to be self-protecting.
And it typically started with data. You know, what do you do to say, “Okay, what are my policies for data? How many copies do I have? What happens if someone gets it, etc?” But the threat matrix has gotten way more difficult and now you have to connect the network and the hosts to the storage, and look at and do a STRIDE review in kind of a security term across that entire vertical application stack.
Patrick Moorhead: Right.
Prakash Darji: So I think we’ve announced a bunch of things here, from secure application workspaces to help secure AI container all the way down, all the way through cyber resiliency capabilities to look at operational configuration, advancements in detection, and just in case something happens, a way to get back up and running with SLAs.
Patrick Moorhead: Well, it makes a lot of sense, because literally, the data is sitting on your devices and in your services. And why not offer cyber resiliency at the point of data? That’s basically …
Daniel Newman: Yeah, I mean, it seems like a really substantial opportunity. And I think what you were alluding to, Prakash, is that with the advent of this AI movement, the vulnerabilities are exponential. It’s not a little bit more, it’s substantial in every different area. And of course, these technologies also enable the black hats and the bad actors to be more sophisticated. You know, every great invention we have tends to get skewed into being used in two ways.
And unfortunately, for the market, that means they’re going to need to invest more. But that’s a good thing in terms of companies stepping up and solving it. It sounds like here at Pure Accelerate, that’s what you’re doing. You’ve made some announcements around cyber resiliency. Talk a little bit about what you announced in particular for the Pure customer base and how they can benefit from what you’re doing in cyber resiliency.
Prakash Darji: Yeah. So the first thing is we do a security assessment. That assesses how well you’re operationally protected. Is your front door open? Have you rotated passwords?
Now this-
Patrick Moorhead: Now is that a service? Is that a service or is it a piece of software that helps you access it?
Prakash Darji: It’s a piece of capability in Pure1 where a customer can see this. We have all this-
Patrick Moorhead: Okay, so we’re not talking about a professional services group has to come in and … Okay.
Prakash Darji: Nope. You can now see when you’re actually doing poorly. And Shawn introduced this idea of a co-pilot-
Shawn Hansen: Yeah.
Prakash Darji: … That actually you can ask, “What do I have to do to fix my access problems?”, or, “How do I benchmark against other people?” But we also then servicized it to say, “Okay. You know what? You could continually do this and you could continually keep opening front doors as someone provisions something. Since we’re running a service and operating the storage on your behalf, why don’t we have a resilience SLA where anything we detect proactively will change the configuration, rotate your passwords, and keep you current and secure, and give your CISO a document telling them all the things we’ve found to continually improve your posture?”
Patrick Moorhead: Yes.
Prakash Darji: So we’ve architected in a way where a customer can take responsibility or they can trust the vendor to take responsibility for remediating these things. On top of that, we’ve announced that we can take a look at new attack patterns. Because as you mentioned, AI in the hands of the bad guys is kind of scary. Now data exfiltration and denial of service are way bigger attack patterns, then I’ve encrypted everything and the data reduction has changed. Right? People have moved beyond, “I’ve just encrypted everything.”
Now they’ll change one block header and be like, “Well, you still can’t access it anyway.” Right? So a lot of the attack patterns have changed, and by looking at latency dynamics in the context of management commands, we can tell what is signal versus noise and identify data exfiltration or denial of service type attacks. And then lastly, in case something happens, we’re there for you. We’ll ship a clean room array with all the PS required to get you back up and running.
Patrick Moorhead: Now, it sounds very holistic. And it’s funny. Years ago, I think I was asking Charlie and company about, “Hey, it makes sense. The data’s there. You have some level of compute that can help you do this. What are you doing there?” He said, “Stay tuned.” And here we are. Now it’s great. You brought up copilot, and Shawn, you announced a copilot. And I mean, copilots are everywhere. And whether it’s horizontal across a data set, we’re seeing vertical copilots across finance, CRM, heck, even ERP and SCM, why do we need a copilot for data storage?
Shawn Hansen: Well, I think it’s an exciting time for us to be in.
Patrick Moorhead: Yeah.
Shawn Hansen: The ability to look across all the billions of data points in a storage platform and to be able to derive new insights really comes down to the kind of questions that you can ask.
Patrick Moorhead: Right.
Shawn Hansen: We think that AI copilot is a radically new way of managing storage than we’ve ever seen before.
Patrick Moorhead: Okay, so you can actually change things through copilot as opposed to maybe querying and it telling you about the environment?
Shawn Hansen: Right. We have this unified operating system for a long time called Purity. Now we have Fusion, which is a unified control plane, and we create hundreds of petabytes of clean metadata. And AI fundamentally is an AI challenge; the ability to see across your entire fleet. I had a brother-in-law who works in the storage space. He’s been administering storage boxes for a long time. And he came to me and asked, “How do I stay up to date with everything that’s happening? This industry is changing so fast. How do I keep up?” Well, AI is this interesting way that we’re allowing people to see what’s happening across all your peers.
Like Prakash just mentioned, now with this AI copilot, you can benchmark yourself against all of your peers. You can say, “How am I doing against all of my peers who are like me? Against a thousand other companies in my cohort, what am I doing well, what am I not doing well?” And your status right now, which might be up-to-date, in six months may be way out of line. So we can look across that entire thing end to end and give you recommendations on how you can level up your game and improve 10% against your peers.
Daniel Newman: It’s funny, I mean-
Patrick Moorhead: Oh, wow. That’s exciting.
Daniel Newman: … Our great audience in The Six Five, they could hit a copilot or perplexity and say, “That great conversation with Shawn and Prakash, and Pat and Dan, give us the summary. We don’t have 20 minutes to watch it.” Actually, Pat and I may have double checked what Charlie talked about doing, but if we watch it, it goes on for an hour. But this is the way we keep up with the world. When information gets created fast, you got to build solutions that enable customers to access and use all the technology you’re building more readily.
Patrick Moorhead: Well, and I love the addition. And as analysts, we like to pretend like we learn it the first time we get the briefing, but comparing against your peers, it adds a ton of value. Obviously, it’s private, you don’t know who others are, but it’s almost an internal benchmarking. And it could have been just a dashboard of something with a bunch of charts, and heck, you could probably make your copilot do that the next version, but it’s another thing to be able to ask it what to do, “Where do I rank on this?”, to be able to query a system. And I have to think too it’s an efficiency play for a data storage manager too. Or am I being a virtual product manager here?
Shawn Hansen: I think so. I think that’s available for every role. This kind of skin will apply to every industry in the end.
Patrick Moorhead: Okay.
Shawn Hansen: But at the end, it comes down to data. I think what you’re going to see in the end is that this will take a decent storage admin and make them good, and it will take a good storage admin and make them amazing.
Patrick Moorhead: Right. That’s cool.
Daniel Newman: It’s exactly the way we see AI being implemented. It doesn’t replace quality, it enhances it, it makes it better.
Patrick Moorhead: No, super powers. It uplifts.
Daniel Newman: All right, we got a couple minutes left. Prakash, just want to hit you up on some of the SLAs you announced here at Accelerate this week. Can you talk a little bit more about those?
Prakash Darji: Well, so the first thing is with Evergreen//One, we found a new workload, the AI one, that didn’t fit the paradigm that we’ve seen in the past. You know, traditionally, with VMware and SQL Server and traditional workloads, even unstructured workloads, more data equals more performance. You need data, you need bandwidth and latency for those applications, but they’re well-defined application types. AI is kind of weird where I think right now everyone’s like, “I need to do something. I don’t know what I want to do yet. I’ll figure it out.” And when you get down to training and infrastructure, everyone’s like, “I can create a reference architecture.”
Right? And we’re just like, “Come on. There’s got to be a better way.” So when we started taking a look at, “Well, GPUs are an expensive asset, and training … ” You know, I’ve even run our own testing in metabolic health. I was actually doing something we can talk about later to learn about LLMs. I realized, “Hey, I can do the training on one terabyte of data.” It’s just generating millions of parameters that are relevant on a small data set, right? And unless you’re a Meta or an Amazon, most of the data sets are sub 50 terabytes in training.
Daniel Newman: Right.
Prakash Darji: And compute-intensive, but when you get to applying the model to inference, it’s a large data set, lower throughput thing. So we’re like, “Okay. If that’s similar to your provision to pipe your home with a water bill … ” You get a one inch, two inch, three inch, four inch pipe, it guarantees the throughput of the water to your home-
Daniel Newman: Right.
Prakash Darji: … And you pay for marginal usage and data. So we had to reorient away from data to solve this AI problem. We said, “It’s a provisioned performance for GPU, is what your reserve commit is, and you pay a marginal data rate like your water bill for using water.” And that concept applies in a way where you don’t know what your requirements are, you don’t know what your apps are, the landscape is changing very drastically. How do you have no sunk cost and flexibility to grow as you need? So the consumption model works well, and our SLA is now guaranteeing the throughput you need to keep your GPUs busy.
Patrick Moorhead: Gosh, imagine paying for only what you need. And your AI needs vary. I think even Charlie mentioned, I think five different use cases that require different types of performance, different types of latency and … No, I love that. It’s good.
Daniel Newman: Well, listen, gentlemen, I want to thank you so much. I know there’s a lot going on here at this week’s Pure Accelerate. I’m sure you’re going to be going from meeting to meeting to meeting if it’s been anything like it has for us. Congratulations on the launch, all the enhancements and Fusion.
Shawn Hansen: Thanks so much.
Daniel Newman: Next big moment. Can’t wait to see how that translates in the numbers. And of course, the cyber resiliency updates, both Pat and I tend to believe this is a very big moment for security. And while AI is going to always kind of have the lead, security, wherever it goes, is going to have to follow. Let’s have you both on soon. Thanks so much for joining us here on The Six Five.
Patrick Moorhead: Thanks so much, guys.
Prakash Darji: Thank you so much.
Shawn Hansen: Thank you.
Patrick Moorhead: All right, everybody. Stay with us. We are here at Pure Accelerate. Join us for all the coverage we have. Check out the links. We’ve had a bunch of different episodes recently with the team at Pure. And of course, join us for all the episodes here on The Six Five. We’re On The Road. It’s Pure Accelerate 2024. We’ll see you all later.
Author Information
Daniel is the CEO of The Futurum Group. Living his life at the intersection of people and technology, Daniel works with the world’s largest technology brands exploring Digital Transformation and how it is influencing the enterprise.
From the leading edge of AI to global technology policy, Daniel makes the connections between business, people and tech that are required for companies to benefit most from their technology investments. Daniel is a top 5 globally ranked industry analyst and his ideas are regularly cited or shared in television appearances by CNBC, Bloomberg, Wall Street Journal and hundreds of other sites around the world.
A 7x Best-Selling Author including his most recent book “Human/Machine.” Daniel is also a Forbes and MarketWatch (Dow Jones) contributor.
An MBA and Former Graduate Adjunct Faculty, Daniel is an Austin Texas transplant after 40 years in Chicago. His speaking takes him around the world each year as he shares his vision of the role technology will play in our future.
