AI raises all boats and the Mainframe turns 60 – Infrastructure Matters: Episode 36

AI raises all boats and the Mainframe turns 60 - Infrastructure Matters: Episode 36

In this episode of Infrastructure Matters, hosts Steven Dickens with Krista Macomber and Camberley Bates cover the a peak at the 60th birthday party for the Mainframe, Rubrik’s IPO and AI tech from AWS to memory and CXL technology.

Key moments from the conversation include:

  • The Mainframe turns sixty, and Steve and Camberley look back and forward on the amazing endurance plus the new adoption of the mainframe.
  • The Infrastructure Matters team lends a view into the recent Amazon’s Just Walk Out technology, and what is evolving within AI and computer vision technology and where to expect Amazon to go next.
  • AI was the top conversation at MemCon 2024 where the importance of memory technologies like HBM and CXL are addressing performance bottlenecks in high-performance computing and AI.
  • How we will see data storage and other technologies gain traction as the new AI gains hold in everyday life.
  • Rubrik’s IPO announcement prompts discussion on its financials, its acquisitions, and the broader trend of data protection companies pivoting towards cybersecurity amidst growing cyber threats.

You can watch the video of our conversation below, and be sure to visit our YouTube Channel and subscribe so you don’t miss an episode.

Listen to the audio here:

Or grab the audio on your favorite audio platform below:

Disclosure: The Futurum Group is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this webcast. The author does not hold any equity positions with any company mentioned in this webcast.

Analysis and opinions expressed herein are specific to the analyst individually and data and other information that might have been provided for validation, not those of The Futurum Group as a whole.


Steven Dickens: Hello and welcome to another episode of Infrastructure Matters. I’m your host for today, Steven Dickens, and I’m joined as always by Krista Macomber and Camberley Bates. Hello and welcome. How are you ladies?

Camberley Bates: Yeah, we’re hanging in there. Ready for another one.

Steven Dickens: Yeah.

Krista Macomber: Ready for another one, for sure.

Steven Dickens: Another packed week in tech, right?

Camberley Bates: Mm-hmm…

Krista Macomber: As always.

Steven Dickens: Oh, as always. So we’ve got a packed show for you today. We’re going to dive straight in here. I spent some time with Amazon and their AI team yesterday. Lots of mainstream media press coverage about some stuff that they’re doing in retail, so I’ll cover that. We’ve got the mainframe 60th birthday coming up this weekend. We’re recording on the 5th of April, so the birthday’s on the 7th, and I think I can do some celebrations. So we’ll do that and we’ll goof around. We’ve got MemCon, and we’ve got some updates around DSPM/DP. I don’t know what that means, but I’m sure I’ll get smarter at some point during this podcast. And then Rubrik’s IPO.

I’ll go first. This week, Bloomberg covered some announcements that Amazon were making about smart carts and their Just Walk Out technology. There was one phrase in that Bloomberg article, one sentence that’s kicked off a whole media storm. I spent some time on a scheduled briefing and then got opportunity to engage with John Jenkins, who also runs Just Walk Out Technology for Amazon this week to really just understand it, because when I saw this break, it just didn’t make sense to me. The key sentence in the Bloomberg article was that Amazon had 1,000 people in India watching live video to be able to augment the AI in the Just Walk Out technology. It didn’t jive with me-

Camberley Bates: Wow.

Steven Dickens: As soon as I read it. That comment’s been picked up and just gone viral. I’ve seen it on social, I’ve seen it in a number of articles. So instantly I reached out to the AR team. Fortunately, I was with Amazon yesterday down in New York on the 4th of April, spending time with them, getting an analyst briefing for the day on AI, so I’ve got opportunities to speak to Dr. Matt Wood, who’s the VP over there for AI. They connected me with John Jenkins as well, who’s the actual VP for Just Walk Out and I’ve got a research note that’ll publish that we’ll put in the show notes. But Amazon is committed to Just Walk Out technology. What they’re seeing is a bifurcation of how people shop, which makes sense.

I spent some time talking to Matt around the small form factor store where people go in the city and I’ve done this in New York. You go in, you want to grab a sandwich, you want a pack of chips, you want a soda. You grab those three things, and you walk out and it’s a fantastic experience. Then obviously Amazon’s got larger format stores. They’ve obviously got the, I’m drawing a blank, Whole Foods as a form factor as well. Those are the longer shopping experiences. People are doing a lot more consideration. They’re walking around, they’re moving around the store differently. So dash carts and Just Walk Out are going to be part of Amazon’s strategy.

I asked Matt specifically about the comments that was in the Bloomberg article. He wouldn’t commit to the size of the team they have in India, but they’re doing training. Key for AI is you’ve got to be able to look at a whole bunch of material, and then you’ve got to be able, especially in computer vision use cases for AI, which this obviously is, you’ve got to be able to go, “Yes, that person put something in their basket. No, that person didn’t put it in their basket.” You’ve got to train the AI models, and in some cases, that’s a people-based training, and I think that’s what’s going on here. It’s been misunderstood by a tech reporter and a bunch of tech reporters in the mainstream press that really don’t understand the mechanics of how you train an AI model.

Camberley Bates: This is interesting. We have been working on a project with Dell on proof of concepts for multiple industry environments, and that has to do with training. One of the ones that we’ve been working on is retail. I think we’re close to getting the paper published. It’s not out yet, but we also worked on manufacturing and digital twins and a bunch of others. I’ve been watching on a weekly call for the last, I don’t know how many months, as we’ve been going through this proof of concept, which is nowhere near about what you’re talking about doing with the AI in the store.

My applause to Amazon, and I’m not the hugest fan of Amazon all the time, so this is a big deal. My husband would know this. My applause to them is recognizing the difference between how we shop, and also the difficulties in doing that. It’s a really super amount of training to do because I’m thinking about when I’m pushing the cart through, I’m taking something out, maybe I put something back in, and all that registration to be able to understand what it is. And then also, when you’ve got it stacked like that, you can’t just walk out the door because, well, maybe you could, but recognizing everything you think that’s in the basket. So really fascinating piece of work.

Steven Dickens: Have you tried a Just Walk Out store? Have you tried one? Because they are fun. It is-

Camberley Bates: I haven’t. I haven’t even been in one. Actually, I have been in one. I was in one that was all kinds of variations of books, like a typical Amazon store with books and gadgets and whatever. It wasn’t the shopping kind of thing. I have not been in one of those.

Steven Dickens: How about you-

Camberley Bates: I’m wondering if that would help with the theft problem. And here’s an interesting piece. I was robbed at a Whole Foods 15 years ago.

Steven Dickens: Oh, sorry to hear that.

Camberley Bates: I just walk in and not even thinking about leaving my purse in the basket. Somebody just came, my back was turned. I was getting whatever, green beans or whatever, somebody came, probably just picked up it, and I didn’t realize it until I was at the checkout because, “Where is it?” And as I went through the process, they said, “Yeah, this is a common place that they target,” because it’s a higher end people or whatever. Just being in that kind of store, your boundaries are down because it’s Whole Foods, it’s protected, it’s nice people, and so your boundaries are down. They do have, especially in Boulder, which is one of the top stores in the nation, they do have the police that are there, or the security people that are always walking around. I’m wondering if that is also going to help address some of that problem that happens.

Steven Dickens: Yeah, for sure. Krista, have you tried a Just Walk Out?

Krista Macomber: I have not, but I love the premise. I am the type of person that if I can go through the self-scan at the grocery store, I am there all day long. I love the premise, but we don’t have one quite near me. Although I will be out in San Francisco next week, so I’ll have to keep my eye out. Maybe there’ll be an opportunity to give it a shot out there.

Steven Dickens: Yeah, I was in the Whole Foods the other day, and it was one of the stores where they’ve launched the palm scanning technology. I don’t know whether you’ve seen this, but again, Amazon retail team innovating. Literally your checkout experience is a palm scan.

Krista Macomber: Wow.

Steven Dickens: So you just put your little scanner on the point of sale, you just put your palm on, bang. It’s doing fingerprint recognition obviously and it was a check-out. I was chatting to the lady while I was checking out how many people, and she’s having a whole bunch of people sign up because it’s that kind of everything’s going on. You’re trying to pack your bag, your purse or your wallet is in the pocket. You’re trying to scramble at the till and there’s people behind you, you’ve maybe got kids with you so just being able to put your hands. So I think some fantastic innovation from the team there.

Camberley Bates: I know you wanted to go it into the mainframe, but I’d like to drop into MemCon because it’s-

Steven Dickens: Yeah, let’s do that first so it’s not me, my subjects.

Camberley Bates: Because it hops over to the AI piece of it we’re talking about here. So MemCon is about memory. Who cares about memory? The AI-

Steven Dickens: We care about it.

Camberley Bates: Really care about memory.

Steven Dickens: Infrastructure Matters would be a really good name for a podcast.

Camberley Bates: Yes, it would be. So MemCon is, it’s a little bit of a small conference. It’s done by a private organization, very well attended by a bunch of tech people. PhDs out the Ying Yang. You have people like Netflix and Facebook I think was there, Meta was there last year. Google’s been there. Some of the lab people like the national labs like Livermore and those folks are there. And the reason, why are they there? Why do they care about memory? Is because memory right now is the bottleneck that we have within high performance computing and AI and we can’t seem to get, what we’ve talked about for many years is this issue about the memory wall, where the memory capabilities of getting memory to CPU has not kept up with the CPU power.

And the same thing as with the GPUs. They are very, very hungry for this in an OAP. So you had, I think it was the guy from Netflix, talked about all the gyrations and methodologies they’re using to get around this problem. So the conversation is about what are we doing in this area? There’s two big technologies they were talking about. One is HBM, high bandwidth memory, and the other one is CXL. Last year we were talking a lot about CXL, which is more or less a switch that it’s like SAM, what we’re doing is pooling memory to be able to feed to multiple CPUs at the same time. So there’s about three different, there’s version one, version two, version three of the CXL capabilities out there. But suffice to say, it’s different to be able to pull memory and be able to feed both the GPU and the CPU market.

What really came out of there is first and foremost is high bandwidth. High bandwidth, what they’re doing is they’re stacking, so today, they’ve got what? Four stacks, they’re going to eight stacks, and there will be problems as we get to higher stacked areas because of the signal times going in between them so they’re wondering how far we can actually go with that before that starts to break down with it. Also, high bandwidth memory tends to be, obviously they’re going to charge more for that, and where the cost benefit is. The other technology that was introduced that was pretty darn fascinating and we are going to dig into this more, is a company called Kove, K-O-V-E. Jonathan Overton talked about this there, it virtualizes memory.

Think about VMware virtualizing CPUs, servers, he’s talking about virtualizing memory and saying this can scale. So there was a lot of skepticism about him. He’s been around for quite some time perfecting this. It appears that there’s some big companies that have been implemented, so expect to hear more from us on this one as we dig into another technology there that can hit into the memory space. So pretty interesting. Samsung was a big, give a call-out to them, they’re a big sponsor there. They are releasing their new CXL. CXL, if we have the triangle for the storage hierarchy or the data hierarchy and you got DRAM, you got HBM.

CXL is that layer between the HBM and that next solid-state drive in terms of speed and access. We’re looking forward to that. It’s not going to address the training issues, which is very obvious. This is why I’m getting back to what you’re doing. It’s not going to address all the trainings that we have, but it will address some of the other applications that are really eaters of memory, and that would be in memory databases in particular is one of the biggest ones that those areas are going to address. So that’s what came out of that group. Samsung is releasing their CXL offerings, expect that to pick up. We didn’t talk as much about CXL because the sexy stuff right now is AI, right?

Steven Dickens: So Camberley, is the right way to think about this is AI is raising all boats? It’s a rising tide, raises all boats, and that’s coming through the GPU, the CPU, the memory, all of the infrastructure stack as I’m not a memory guy, but I can see that that would be what you are saying. Is that the simplest-

Camberley Bates: Did you read my post this morning or something?

Steven Dickens: Oh, of course. It was the first thing I did when I woke up.

Camberley Bates: Okay. So yeah, my post this morning is I reposted something that Daniel was on the news today, I think it was Bloomberg or CNBC or something like that talking about semiconductors. What I was saying is that here is the pattern of what’s going to raise all boats. You got GPU to CPU to memory, to networking and to data and the data storage, and the data storage is the last one I’ll make. That was the other point that I was going to make today was what we’ve seen just recently, literally the last 30 days or so, and I believe it was Samsung as well that talked about this and published about this, about the demand for their solid-state drives has kicked up and significantly kicked up. And also, what we’re going to see is a rise in prices just like we’ve seen a rise in price in memory, we’re going to see a rise in price in solid-state drives. Is that AI? Maybe. Is that an energy efficiency problem where we’re having to swap out hard drives for a solid-state? Maybe. More to see.

Steven Dickens: Fantastic. We’ll go to you next, Krista. Rubrik announced their IPO this week or announced that they’re planning to IPO with their S-1. I saw that picked up.

Krista Macomber: It did.

Steven Dickens: I didn’t get to track it, but that’s why we have a fantastic team on the, so tell us about Rubrik’s week.

Krista Macomber: Yep. Yeah, 100%. This is something that we’ve been as an industry expecting for some time now, so it was interesting to get the actual filing. There was some buzz around the net losses that Rubrik has been reporting. So they had 354 million for their year-end, that was back at the end of January of this year. That was actually an increase from just under 280 million in the year before, so I saw a little bit of buzz around that. Just when we think about Rubrik in the context of cybersecurity companies, certainly it’s a large net loss. Now that being said, certainly not unexpected considering the investment that not only Rubrik, but all of its peers really in the data protection and data management space are having to make to broaden and evolve their capabilities to address things like growing cyber threats.

So that is, for example, Rubrik made an acquisition of a company called Laminar towards the end of last summer. They provide what we call data security posture management, which effectively is going beyond reverse core capabilities in areas like data discovery and classification to be able to actually, for example, identify misconfigurations, other threats and vulnerabilities so that way IT and security teams can address them. So long-winded way of saying it gave us a little bit more insight. They have not yet actually filed. We’re thinking maybe that will happen at some point next week.

But just another interesting comment is that when we think about data protection companies, it’s been, I can’t even think of the last time that a data protection company went IPO. Cohesity was potentially a contender, but now they’re acquiring assets from Veritas so we would expect for that deal to go through before potentially if that combined company did choose to go public. It’s another component that makes the announcement interesting, and I would really say just underscores the focus and the emphasis on cyber resiliency that we’re having today.

Steven Dickens: I mean, is this, the likes of Rubrik, Veeam, Cohesity just becoming hot because of the pivot from data protection to cyber securities? I mean, that’s some of this pivot I see going on. Data protection was back up. We’ve been doing that for years. I see the whole sector moving more to a security perspective. Is that a right way to simplify it?

Krista Macomber: I think it is, and in fact, just yesterday I recorded a fun podcast with our Tech Field Day team because we’re all going to be at Security Field Day next week. The premise of the podcast was cyber resiliency is just data protection. And of course, that is not true. It was our premise to just start some discussion and get some eyeballs on the podcast. But we were talking about the fact that data protection plays an important role in being resilient against cyber attacks.

But when we think about data backup, data recovery replication, it goes much further beyond that. I talked about data security, posture management, and the ability to be more proactive and identify threats so that hopefully you can prevent an attack from occurring. We also have capabilities being introduced into data protection software that can detect anomalous user behavior or file activity that might indicate that there is an attack that’s going on. We are definitely seeing that data protection plays an important role, but again, it’s certainly going beyond strictly the recoverability of the data itself.

Camberley Bates: We had the conversation yesterday, Krista, about the speed of recoverability, and the issue about how fast can you get back up and running. That has a lot of different implications if anybody read the recent article on MGM and how what they went through and that entire fiasco that happened over there. But if you really start looking at what the CEO, they ended up doing, they didn’t pay money out. Whereas other people in Vegas did pay money out on this, at least didn’t pay money to the ransomware people. They paid other people, in terms of the gambling.

But the big thing is taking those servers, having to completely rebuild the servers all the way back up. Then making sure that you’ve got a known good place to come back in, and where is that known good place to come back in and how fast can you come back in? So it’s a village of IT people to get things back up. It’s not one discipline. There’s no single discipline and that gets back to what you were saying is that the reason why we’ve seen this uptick is because of cybersecurity, but because all of those teams have to come together and work well together to bring a company back to status.

Krista Macomber: They really do. Just another comment on the recoverability, absolutely speed, but also having the insight and the context into not only what systems and data were impacted, but what ones are critical to business services, and maybe which data is most sensitive that has been impacted, right? Because it’s that way we can prioritize the recovery processes and be more informed from that perspective.

Steven Dickens: So the final topic here, and one dear to my heart is mainframe 60th Anniversary. I’ll do my thumbs up to get the celebration going. We’ve recorded with a bunch of the senior leaders, a whole bunch of thought leadership videos that talk about the platform and where it’s going. That’ll air on Monday, the 8th of April. Don’t know when you’re watching this show, but you’ll be able to check that out if you just Google for Mainframe Mini Summit and Futurum, you’ll find us there. But really just an opportunity to take a moment, reflect on this platform and what’s been going on.

The main frame, one of those foundational pieces of technology and probably 5,000 plus enterprises globally now, foundational technology that supports our credit cards. It supports retail, flight bookings, a lot of your banks, the IRS uses it. A lot of tax departments around the world use this platform. Really huge platform that’s been able to continue to innovate, continue to grow, and evolve. The Telum processor was the first to get to seven nanometers. It’s got AI acceleration, cryptographic processing, quantum safe, cryptography. We’re not talking about a historical platform here. We’re talking about a platform that’s really able to continue to be on the bleeding edge of innovation. I’ll pause because as anybody knows, I could talk about this for hours.

Camberley Bates: I was reflecting on that, because I started my career with the mainframe and I’m older than you, so there you go.

Steven Dickens: Our listeners and viewers won’t believe that to be true, but it technically is.

Camberley Bates: It is. I started my career with mainframe in IBM in the mainframe division, so I have a warm heart for that. But as I was thinking about 60 years and what it survived. Sun Microsystems, they’re still out there but they’re not being sold. You had DEC out there but not being sold. IBM System/36, no longer. IBM Series/1, I don’t think it’s out there anymore. You look at some of the other minis, computers or other computers that were developed. Burroughs, Unisys, Amdahl, all of those are gone. They’re all gone. And what we have now, all of those were replaced by x86 or whatever, and customers did say the mainframe is dead, all that kind of stuff in the 90s and even into the 2000s, we basically said that. Customers were diligently trying to get off that system, and yet today that box is growing in terms of size.

Steven Dickens: And they said 16 cycles, the best they’ve had in decades.

Camberley Bates: You got to ask yourself why? Is it because it was so hard to get off, they couldn’t possibly get off, but they could get off the others? So that to me, that doesn’t hold about why it’s still there. I believe the reason why it’s still there is because it is such a hardened system. In terms of when you look at IBM research, the first research capabilities that come out usually come out on the mainframe, and then they trickle down into some of the other stuff that IBM has, whether it’s the storage or there’s, I just took a brain-

Steven Dickens: Power.

Camberley Bates: Power, thank you. Their power. I’m going risk, this one is like, geez, Camberley, you’re so far back. Trickles back into those things. So it is a testament to that crew. I know you were involved with the LinuxONE, so that actually took it to a new world. And then they’ve also been heavily investing in the people and the people’s skills and bringing in Gen Z and millennials and all that kind of stuff to be able to, as we see these people retire, i.e. Randy Kurtz, 50 years in, he’s done. He’s not doing that anymore. So a next generation has to come in. So I’ll stop my waxing on about my love for the mainframes.

Steven Dickens: Yeah, this is a love in here. But no, I don’t know when this is going to air, probably after the event. Now we record this on Fridays and it publishes the week after. But I think certainly check out our content. There’s plenty to be looking at there. There’s probably I think two or three hours worth of chatting to thought leaders about where the platform’s going. We’re not taking that retrospective look back.

Camberley Bates: Okay.

Steven Dickens: We are certainly taking a forward-looking view of the significance of the platform and where it fits so recommend you check that out, we’ll put that in the show notes. Packed show this week. As always, pleasure to talk to you both. You’ve been watching the Infrastructure Matters podcast where every week we talk about all the big news and insights that are coming from the world of technology. Please click and subscribe and do all those things to make sure this show continues to be one of the highest ranked around the technology space. Thank you to all our listeners, and we’ll see you next time. Thank you very much for watching.

Other Insights from The Futurum Group:

Futurum Group: The Mainframe Summit

Amazon Under Fire: Just Walk Out Technology Causing Waves

Exploring Samsung’s AI Ambitions – The Six Five on AI

Author Information

Camberley brings over 25 years of executive experience leading sales and marketing teams at Fortune 500 firms. Before joining The Futurum Group, she led the Evaluator Group, an information technology analyst firm as Managing Director.

Her career has spanned all elements of sales and marketing including a 360-degree view of addressing challenges and delivering solutions was achieved from crossing the boundary of sales and channel engagement with large enterprise vendors and her own 100-person IT services firm.

Camberley has provided Global 250 startups with go-to-market strategies, creating a new market category “MAID” as Vice President of Marketing at COPAN and led a worldwide marketing team including channels as a VP at VERITAS. At GE Access, a $2B distribution company, she served as VP of a new division and succeeded in growing the company from $14 to $500 million and built a successful 100-person IT services firm. Camberley began her career at IBM in sales and management.

She holds a Bachelor of Science in International Business from California State University – Long Beach and executive certificates from Wellesley and Wharton School of Business.

With a focus on data security, protection, and management, Krista has a particular focus on how these strategies play out in multi-cloud environments. She brings approximately a decade of experience providing research and advisory services and creating thought leadership content, with a focus on IT infrastructure and data management and protection. Her vantage point spans technology and vendor portfolio developments; customer buying behavior trends; and vendor ecosystems, go-to-market positioning, and business models. Her work has appeared in major publications including eWeek, TechTarget and The Register.

Prior to joining The Futurum Group, Krista led the data center practice for Evaluator Group and the data center practice of analyst firm Technology Business Research. She also created articles, product analyses, and blogs on all things storage and data protection and management for analyst firm Storage Switzerland and led market intelligence initiatives for media company TechTarget.

Krista holds a Bachelor of Arts in English Journalism with a minor in Business Administration from the University of New Hampshire.

Regarded as a luminary at the intersection of technology and business transformation, Steven Dickens is the Vice President and Practice Leader for Hybrid Cloud, Infrastructure, and Operations at The Futurum Group. With a distinguished track record as a Forbes contributor and a ranking among the Top 10 Analysts by ARInsights, Steven's unique vantage point enables him to chart the nexus between emergent technologies and disruptive innovation, offering unparalleled insights for global enterprises.

Steven's expertise spans a broad spectrum of technologies that drive modern enterprises. Notable among these are open source, hybrid cloud, mission-critical infrastructure, cryptocurrencies, blockchain, and FinTech innovation. His work is foundational in aligning the strategic imperatives of C-suite executives with the practical needs of end users and technology practitioners, serving as a catalyst for optimizing the return on technology investments.

Over the years, Steven has been an integral part of industry behemoths including Broadcom, Hewlett Packard Enterprise (HPE), and IBM. His exceptional ability to pioneer multi-hundred-million-dollar products and to lead global sales teams with revenues in the same echelon has consistently demonstrated his capability for high-impact leadership.

Steven serves as a thought leader in various technology consortiums. He was a founding board member and former Chairperson of the Open Mainframe Project, under the aegis of the Linux Foundation. His role as a Board Advisor continues to shape the advocacy for open source implementations of mainframe technologies.


Latest Insights:

Kate Woolley, General Manager, Ecosystem at IBM, joins Daniel Newman and Patrick Moorhead on Six Five On The Road to share her insights on the growth of IBM's Partner Plus program and the strategic importance of partnerships in the AI landscape.
Dr. Darío Gil and Rob Thomas from IBM join Daniel Newman and Patrick Moorhead on the Six Five On The Road to share their insights on how IBM's AI initiatives are driving significant transformations and value for enterprises across the globe.
Tina Tarquinio, VP at IBM, joins Steven Dickens to share her insights on leveraging AI with the mainframe to enhance productivity and modernize applications, charting the course for future innovations in IT infrastructure.
New Catchpoint Capability Transforms Internet Performance Monitoring with Its Real-Time, Comprehensive Internet Stack Visualization
Paul Nashawaty, Practice Lead, and Sam Holschuh, Analyst, at The Futurum Group share their insight on how Catchpoint's Internet Stack Map affects IPM by enhancing real-time, comprehensive monitoring capabilities.