RSA, Red Hat Summit, AWS and Solidigm – Blockbuster Week in Tech – Infrastructure Matters, Episode 41

Hello and welcome. My name's Steven Dickens, and you're joining us here on what is now episode 41 of Infrastructure Matters. I'm joined as always by Krista Mako and Kimberly Bates, welcome to the show.

Thank you. And we missed you last week. We did, We did.

I was in the UK and not really podcast ready, so just like you guys probably did a better job without me. We, we told everybody, we told you they were, you were over there with family, but also doing work. So there you go.

Yeah, they had me doing some work. I did, uh, got time to spend with Rocket on their announcement around their closing of the A MC acquisition from what is now OpenText. So yeah, you'll see that video's come out on our channel, so check those out.

So I think we've had a quiet week. I think we've all just been at home. We've not been on the road.

I'm joking, obviously, I dunno about you, but I got nine, nine solid hours of sleep every single night. I know you did Kristen, right? You were just, you were just cruising through the week right too.

Chris, you, you've, you've had the busiest week of all of us I looked at. We were pinging on, uh, what did you do? 25 vendor meetings this week?

I think it was 25 meetings over the span of 72 hours that I was in San Francisco at RSA. So I have a new goal to beat for next year. I, I, that, that's gotta win an award or something.

There's gonna be like a plaque or Something ever. Did you get any food? You know, we board listening to us, so we probably should get to the meat of this stuff.

Right. So, so I mean, that's a great segue. You've had a crazy busy week, largest security show of the year.

Let's give you the floor. Tell us what you saw. Tell us what was going on.

Yeah, yeah. So, um, it was, so as I mentioned, um, so I was there for, um, for, for three days and, um, I think I met with about 25 companies. So we kicked it off.

Um, every year Microsoft has their pre-day, um, on Sunday afternoon. So we had the opportunity to attend that. Um, well Townsend on the more insight side, um, and myself.

And then we filmed, um, a couple, another kind of great videos, um, with them first thing Monday morning. Um, and then from there it was sort of off to the races. So, um, a great mix of, um, you know, kind of analyst meetings with, um, both companies that we've worked with for quite some time.

Um, some kind of companies that are newer to us. Um, and we filmed some other videos. Um, we got to spend some time with the Textron team, um, who are now part of the, the Futurum group family.

Um, so it was myself, she Rubinoff, will Townsend, um, and the entire Techron team. Um, yeah, we were all there kind of, you know, trying to spread the wealth and, and really cover as much ground as we could. So last week you talked, we, I asked you the question about what were gonna be the big themes here.

Yes. And I, we said ai. So what were the big themes coming out of RSA?

Yeah, so Did anybody talk about ai? I'm just be shocked if they did, Right? No, it didn't even come up.

No, I'm joking. I'm joking. Um, yeah, so, you know, I know for Steven, who I know you were on the road last week and, and for anyone who wasn't able to listen.

So kind of that was obviously, you know, one of my, one of my predictions and I would say it, it did hold true. Um, so kind of what I was discussing last week was, um, and a number of kind of the pre-conference calls that I was having, my, excuse me, my sense was that obviously it's, it's a buzzword, but we're really getting down to, um, kind of some more concrete technologies services, best practices for being able to protect our artificial intelligence applications. Um, so from my side, coming from, you know, kind of a historical focus for me personally on data security and data protection, I'd already been getting a lot of questions around sort of, um, protecting, you know, the data itself that's being used.

Um, primarily that's being used to train artificial intelligence applications. Um, you know, but also we're, we're even getting other tools. So we're even hearing, um, AI security, posture management, which is gonna be another theme that I'm gonna talk about is kind of our, our security posture management.

Um, so, you know, I bring this up to kind of articulate that we're getting some technology capabilities, some services and starting to maybe getting some kind of frameworks to be able to protect kind of the life cycle of our artificial intelligence, um, applications. So I have to give a call out to probably my most interesting meeting of the week, which was with Calypso AI in Waymo, artificial intelligence, uh, autonomous, I should say, um, autonomous driving car. So we kind of were able to take a scoot over to, um, you know, the Fisherman's Wharf area and then back to the Moscone Center.

So it was, um, you know, literally quite a trip, um, especially watching the Waymo kind of navigate how to pull over and let me out in front of Moscone South when there's people in cars and everything like that. But anyway, um, it was kind of a, a good, you know, parallel to, you know, we're starting to be able to trust artificial intelligence, um, a little bit more, or at least we need to make sure that as we're beginning to adopt ai, that we do so in a way that we can really trust it. Um, so for Calypso, what they do is, um, they actually look at the large language models themselves that are being, um, that are feeding AI applications.

Um, and they do things like identify vulnerabilities, for example, and kind of try to identify risks both internally and externally. They do a lot more. But, um, again, that was just kind of one, um, one example.

And, um, if I were to try to summarize all of the variety of app of conversations I had around ai, that would be sort of the way I would summarize it in terms of where we're at from a security standpoint. Okay. So any of the big takeaways, obviously ai Yeah.

Yep. Been a bit of consolidation, m and a activity. Yeah.

Was there any kind of other sort of big trends that we should be thinking about? Yeah, yeah. So for sure I would say, um, you know, the other couple of technology trends, so I kind of just threw a little bit of a reference to this concept of, you know, a threat posture management.

So whether that be, um, we've talked on this podcast about data security, posture management, um, you know, but also we have, you know, really kind of the ability to, as I was referencing, you know, kind of Calypso does this a bit from an AI large language model perspective. Um, I got to meet with snyk, um, and what they do is they try to look at the, um, kind of the code itself that developers are using to create, obviously not just AI applications, but any application, um, and to try to look at kind of the, um, the integrity and kind of the security posture of that code. So this is something that it's in my mind, becoming a little bit of a buzzword.

Um, you know, but at the same time, it's, I would say it's certainly for good reason because we're trying to become more proactive against the attackers and the malicious actors and really kind of become more preventative as much as we can. Um, we're never going to be fully one step ahead of them, but, you know, the more that we can have visibility across our infrastructure, our data, our software code, um, you know, and kind of really under understand where those vulnerabilities are and how they're evolving over time. Um, and then also I'll make this comment, which is gonna lead me into another topic that I wanted to touch on, but, um, ideally gaining some insight that's actionable for, um, you know, the security team, the IT team, things of that nature.

Really try to kind of sift through the noise as well. And I, and underscore where are their priorities in terms of these vulnerabilities to address and what can we do to address them as well? Um, and that, you know, that kind of leads me into, um, another topic, which is just the, you know, kind of the burnout of, you know, security and in IT teams, you know, they're just, we're trying to scale so much across these, um, you know, across, um, you know, whether it be kind of the infrastructure environment on the IT side, but for security teams, um, you know, they're trying to keep pace with, you know, how all of these threat factors are are evolving.

So one of the videos that I filmed was with Elastic, um, and what they're doing is they're, um, essentially they're trying to, um, use artificial intelligence, um, but to provide, you know, analytics around, um, you know, kind of security threats and vulnerabilities and almost offer an alternative to kind of your traditional SIM type tool and offering for the Security Operations Center. Um, so that was a great video. It, it should be live by the time this webinar is live, so we'll make sure to link to that.

Um, but that was kind of another theme, which was, you know, how do we really help to overcome, you know, some of this burnout, help these security teams in particular to scale better, um, and kind of use not only AI, but automation in different technologies to be able to do so. Now, of course, it's, especially when you start to think about using AI for defense and security, it's a little bit of kind of crawl, walk, run. We have to learn to trust it, right?

So typically, you know, it's not necessarily, in my opinion, completely gonna take over the security operations center, you know, anytime soon. But it's, you know, the more we can kind of, um, again, find different ways in our daily processes to integrate it, learn to trust it, and then from there kind of broaden our use. I, you know, my sense is sort of that's where we're at, but that was kind of another big theme that least in my conversations from the week.

Oh, fantastic. I think we could spend the next two hours on those 25 meetings, but we, Kimberly and I had a busy week as well, so, Mm-Hmm. I mean, I'm assuming there's a bunch of research notes coming from you.

Yes. Yep. Cover all of this, a bunch of videos.

Mm-Hmm. So we'll link to those in the show notes. Absolutely.

Before, I got to spend some time with you this week at Red Hat. You had a busy Monday before you joined us. Yeah.

Kimberly, you were at Solid, not a vendor. I know particularly well, I know you know them really well. Do you wanna just maybe position them and then talk about what, because you came, I met you, when was it?

Tuesday morning you were buzzing about your time at Solid. So It was a, it was a great time with the, the, the entire executive team that's there and then some, I mean, they showed up, all of them showed up. I think the only person that wasn't there was the CFO, but I, the entire crew Back in the room counting the dollars.

I dunno, I dunno. But, you know, uh, uh, Daniel, um, was there as well as Patrick Moorehead. We were doing some, some work with them, but we also took advantage of this full day of doing analyst briefings and then going, doing a lab tour, of which I had the horrible mistake of wearing high heels that day.

And it's a really big lap. So I thought it was safe, but I wasn't, I shoulda have worn sneakers, by the way, Kimberly sneakers all week for me, sneakers all week. Anyway, so, but it was Dave, every voice.

I'll do a couple of call outs of some of the names. I mean, just to give you an idea, Dave Dixon, the CEO, um, the Greg Matson, who runs strategic planning and all the marketing, um, Robbie Frick Frickey is the in, he's a fellow with them. Um, they had, they had there as well as a bunch of the marketing team was there.

So I, and then da, um, Dale, uh, Marty Shaw, I think is the guy that's runs the lab, um, that took us through the lab and it was just, it was a great, great trip. So a couple things, if you don't know who Soine is, they are the company that came out of Intel. That was when Intel moved out there, all their solid state drives and the kick, there's all kinds of people there.

You know, there's other people that are doing solid state drives. But the one thing that's unique about solid time is they're focused on one thing and one thing only in that is enterprise solid state drives. And so they're very tied into not only the big hyperscalers and those folks that are looking at each, those kind of drives, but also all the systems people which have now, you know, we know that probably 70% of the data is still on premise that they're supporting.

And so what, what are the unique things that have to be developed and delivered in terms of this, both the AI environment, which is what we're working on. They showed us a whole lot of data, um, in terms of, I don't have that in front of me, um, right now, but they showed us a whole lot of data on terms of performance and capabilities they're doing. And yeah, there's one, one way to look at data, which is a four block that fi IO, um, testing.

But the real work is on the real workloads. And what they find is that when you get into really work, real workloads, which is I think is the area that we're gonna start working with them on the testing that has to be done, is that their, their drives are doing extremely good performance in those environments. So, you know, whether or not we're talking inference training, um, whether or not inference or training or, or some of those areas that are in ai.

So that's what they're focused on. Um, so look for more coming out from us. Yeah, you definitely came away buzzing from that day when I met up with you on Tuesday.

But before we get to Red Hat Summit, one other thing that I wanted to touch on just briefly, AWS investing $9 billion in Singapore and building out their capabilities there. I'm just seeing a trend now. I've written about this three or four times with Sovereign Cloud, IBM investing in Canada, AWS and, and Oracle investing in their Japan operations.

Big theme I see just kind of arc throughout the, um, cloud space, sovereign clouds, it's comes up in pretty much all of the briefings that I'm having that they're looking to build out this capability. The world's a dynamic place. There's lots going on.

5 million population country, and have their thinking through Sovereign Cloud and they're thinking through regulations. So just a real big sort of trend line. You know, AWS investing in cloud infrastructure is not big, big news.

But when you join the dots up around this comes on the back of them investing in Japan back in, um, January, you know, you look at what IBM's done with Canada, you look at what OC i's done also in Japan, really starting to see a trend line around sovereign cloud. And this is because latency's obviously a factor when you're talking about places like Singapore and Japan, but it's also the data's gotta be resident within the country. Well, That's one the problem with the latency as well, right?

Is the data set by the, by the computers. Nobody's fixed the speed of light yet. There's some people working on it, apparently, but nobody's made bits go down fiber quicker.

Um, so I mean, all joking aside, just a big trend line, as I say, not a not a rockstar announcement on its own. I'm gonna cover it and there'll be a research note on it, but a bigger, broader trend that I'm seeing. But let's get to Red Hat.

Kimberly and I spent some time together this week at Red Hat Action packed announcements. We got to spend a whole bunch of time with the CEO, Matt Hicks and Aade Badani, who runs their, um, pro Chief Product Officer. I got to spend some time with Satish Bala Krishnan, who runs their Ansible business.

Fascinating discussions. But Kimberly, what were your key takeaways and thoughts from Red Hat Summit? A a couple things.

Um, Matt Hicks started out by TA Hick making this statement, and I thought it was very curious and then kind of going, okay, so where is he going with this? And that, um, talking about this explosion that we're having in, in the AI and everything else, and there's an explosion of options as well. And then he made the statement, he said, what's in the center of all this innovation?

He said, open source and academia. And I can almost say, yeah, open source is in the, is in the middle of this when we're looking at the, you know, open AI and all things that are going with LLMs and sort, and also the fact that this, most of this is being built on a, in a Kubernetes platform, but why the academia? And he said, you know, the academia are kind of where all these concepts came from, right?

You know, in terms of neural networks and, and the, you know, whether it's IBM research or somebody else's research or, or any of the, the labs that we have. So yes, a lot of this AI work, and that is where it's come from. It's like, okay, so that's where this come from.

And then he went on to say that, and I really appreciate that AI will not be built by a single vendor. It's not a single monolithic model. Um, you have to be able, and you have to be able to run anywhere and will be based, and for, you know, for them they're saying it will be based on open source.

And so what we heard was two big announcements that I came out r AI and which is new. Um, OpenShift AI is not new, but the r AI is new. And what is new about that is the ability to run on your PC or your a i pc.

We just got one in the lab here, you know, in A IPC, I'm gonna run re I'm gonna run AI and I'm gonna be able to do some training. And what did IBM do this week, Steve? I could source granite.

So yeah, I mean that was the one that stood out for me. I was chatting to a reporter there, and maybe they'll put some of my comments in there, article, but I was chatting to a reporter in the break, 'cause Red Hat put the press and analysts in a lot of these sessions for efficiency and maybe being a former IBM or maybe being a former, I, Emory worked on Linux when he was at, at IBMI, I know the context. I, IBM's been all in on open source since 20, the year 2000, they made an announcement in year 2000 that Linux would run on all of its server platforms.

Now, obviously that was the mainframe, but it was also power. It was always also System X. At the time that was pr, that doesn't sound revolutionary now 24 years later, but Linux was six years old as an operating system at the point when they did that.

So being able for a, what was at the time, probably the biggest, what if, you know, certainly one of the biggest server vendors by volume, obviously they've sold the system X business off to Lenovo now. But if you wind back 24 years, IBM saying that Linux would run on all of its systems was a big deal. You know, the GM that I used to work for, Ross Morey used to be part of the, uh, used to be the chairman of the Linux Foundation before it was the Linux Foundation, so Open System Development Lab, OSDL.

So IBM's got a long track record of open sourcing things. Now yes, they've got some proprietary platforms as well, but they've done a really good job over the years. So I think this is on message with IBM, you know, it, it's not surprising to me and it shouldn't be surprising to anybody else that IBM deciding to take something that it's developed in-House and then open source, it is on message for those guys.

So I think, uh, my personal perspective for Granite to get adoption, it's got to be on hugging face. It's got to be open, it's gotta be available as a tool that lots of people can download. So I think kind of they had to do this, but I think it's good that they did, they flashed up some performance benchmarks, not a lot of explanation, something for us to dig into with our labs.

I took a picture of it and tweeted it, but you know, I think they looked impressive. But the devil's always in the detail in the six point font, uh, below a benchmark and needs 400 questions to know whether it was significant. But they were putting up LAMA two and LAMA three, and I need to go back and look, but there was, I recognized all the names.

They blew through it pretty quick. So, but the performance benchmarks looked impressive. So I think it's a meritocracy on hug, gr hugging face.

You put a model out there and you'll know whether it's going to sort of resonate. It's gonna be interesting to see that. I think what, uh, red Hat's doing with granite and building that in is also interesting.

So it's gonna be fascinating to watch that space play out. Well, and that brings me to that next big announcement was, was Instruct Lab. Mm-Hmm.

Which I am not a programmer. So I went over there and I was like, okay, I, I don't know YAML code, so I'm not gonna be able to do this. But Instruct Lab is, so the premise here is that the ability to contribute to a model has yet to be solved.

In other words, how do I, as in, in open source, I'm always looking at how do I contribute to the projects, right? And so they're looking and saying, how can we do this? And Instruct Lab is that vehicle, it's for people to do some training and then to contribute back into these models.

And so, you know, going with their open source kind of concepts saying how do we make this happen? You know, and all the processes that are around there. Um, it's these, there's, um, and I, I think that, you know, it's a brilliant move.

Um, the, uh, the, you know, we saw huge, I mean, you couldn't even get near kind of the, where the, the labs and the, it's Crazy. The lab, I mean, BC conference in general, I thought, but in STR lab it was like standing room only. All the red Hatters doing the demos were just busy the entire time.

So this was, you know, and, and as they were saying is when I sat down and talked to, you know, some of the folks like Will McGrath, who's on the marketing side, um, talking about, you know, how do you create, you know, minimize the barrier to entry for adoption and an open source is very easy. I go to GitLab GitHub and I download something I'm running. Um, it's not supported.

It may be kind of screwy or whatever, but at least ongoing. And so their vision is can we do the same thing with the tools and can we do the same thing with contributing and putting the information back into the LLM? So those are really, really big announcements, um, that, that were going on there.

And, and we've got think, what is it not next week, the week after. I'm expecting the IBM side of all this to be sort of getting to. So I think as you look ahead a couple of weeks to IBM think, obviously Watsonx is gonna get a lion's share of the sort of announcement time, but I think, you know, the open sourcing of granite, what they're doing with Red Hat, red Hat as a big presence at at Think as well.

So it's gonna be fascinating. And the other part of this conference that I really enjoyed, and I guess I didn't appreciate it quite as much last year when she presented, and this is Stephanie Chias, who is Mm-Hmm. I know you think she'd be the next CEO of IBM or something like that.

She's Definitely in the, I I think she's gotta be on the short list. She's just Impressive. She's, her presence is phenomenal.

So yeah, she's good. But, but you know, when I first, you know, heard them talk about, heard her talk about ecosystem, it just sounded like partners. But that's part of the, I think that's part of it because she was just getting her hands, um, talking about other ISVs or hardware people and that kinda stuff.

I think just because she was just getting her hands around what this really looks like and where it is and, and what was very clear from her presentation is this is a very, a methodical thought process Mm-Hmm. Of what she's looking at and how she's building out. And a very, you know, succinct way, this overall program understanding, we have hardware vendors, there are ISVs, there are systems integrators, and then there are the, the regular reseller partners and the distributors.

And she's, you know, very holistic look and view of how will Red Hat leverage all of these folks. And they are, and what was really clear is how important they are to their, the ecosystem and, and managing it. So, and the other piece I saw, you know, they brought in a whole lot more probably best practices from IBM on partnership partners marketing and partner support.

Um, so I think that it's a really, if the partners haven't looked at this yet, they really do need to take a hard look at what Red Hat is doing and given where we're going with all of this, um, embrace it. Well, I mean the, uh, everything you said plus one, I think the other interesting piece for me is Stephanie's done a great job of growing and making partners more strategic. Mm-Hmm.

Whilst the ownership of the company has changed. So there could have been a blip with Red Hat changing its ownership to being owned by IBM, you know, that could have put them in conflict with Dell or ib, uh, with, you know, HPE or Lenovo, or it could have put them in conflict with other delivery partners. GSIs absolutely not the case.

Mm-Hmm. That you see that getting better despite the backdrop. And that's what's impressive.

I mean, I'm, I'm a Stephanie Shiraz fan, been tracking her career for a while. I think she's gotta be in the mix for a bigger job, but e either at Red Hat or in, in the wider IBM, she's, uh, she's definitely doing a great job of that partner program. So the other one that stood out for me, and you've not covered it, was where the Ansible guys are doing, putting light speed across the entire Red Hat portfolio portfolio.

It's been that. Okay, so what is Lightspeed Lightspeed's good stuff around. Just buy it.

Okay. Got, just Buy it. It's good stuff.

Basically a copilot basically is when, Yeah. Yeah. So I mean, I got time to spend time with Satish Ball Krishnan after the, after the Ansible keynote.

I think the Lightspeed everywhere is good. The thing for me is the trajectory IBM's and Red Hat are gonna be on with the Hashi Corp, kind of Terraform and Ansible. I asked some questions, didn't get any answers, but I think you, if you read between the lines, can be interesting where they take that over time.

Obviously that HashiCorp acquisition has gotta close first before they can get into the sort of product management piece. But really fascinating conversation with Satish talking about kind of Ansible as sort of automation as code. Really, really good demo as part of the Ansible keynote.

Talking about kind of how they're applying AI to the whole sort of Ansible mm-hmm. Event driven automation. So I think we're so early and event driven and automation driven, you know, with ai, but I saw the, I saw a trend line, I saw I, this makes sense.

These teams are gonna be asked to do more and more. We've got a wa we, we talked about it off camera, I know ly during this week a wave of applications coming. You know, there's gonna be hundreds more AI applications.

So if you've got an application portfolio of, you know, 500, 600 applications today in an enterprise, I don't think that's gonna double, but I could see it growing by 50% over the next sort of 18 months as all these AI applications come in. The teams are just barely keeping up as it is. They're gonna have to do more in automation just to stand still, let alone and move forward.

So I think automation as code, kind of what they're doing and infusing that with ai, as I say, Lightspeed going everywhere as an a co-pilot. I, I think it's gonna make a lot of sense. So, I mean, we didn't cover it Ansible very much, but I mean, it's been a packed week.

I actually have one more to cover. Can I do one? Okay, go for It.

We'll go over 30 minutes here. Um, and that was vSphere. mm-hmm mm-Hmm.

It was a major point of topic on the show floor. Um, in terms of clients, um, talking what options do we have or are there options within OpenShift? And, and as I think, you know, OpenShift from a Q vert is not necessarily, is not a re total replacement for a vSphere at all.

Plus the conversion is big. But when you ask Matt, Matt Hicks again, you know, where are you guys going with this? You know, because frankly, to their credit, that word vSphere, VMware never got said by anybody on stage.

Um, so, and they might talk about virtualization, but in terms of what they were looking and, and how his answer was, it's very forward thinking. It's saying, we're not going back in time. You know, where we're going.

And what he says is what we're, we're looking at what we did with OpenStack to go forward. You know, we don't wanna be back, we don't wanna look back, we wanna look forward and where we need to take OpenShift capability to be that durable system that can support the full platform. Um, and that there will be different ecosystems.

And so, you know, looking at where they're going and I thought it was a very thoughtful, um, kind considered answer, um, given the hullabaloo that's going on right now. So, Yeah, I'd agree. I thought, I, I thought Matt Hicks impressed me this week.

I've gotta admit he's kind of very visionary, very thoughtful about this stuff. So he still looks Like a kid though. Looks like a kid who's in his forties.

He looks like he's in his forties and he looks like a kid. But we digress. I mean, we've running a couple of minutes long, massively busy week.

I think we could have done the Krista show this week with everything from RSA. But thank you very much for joining us again. We'll keep doing this.

We're up episode 41. We love doing this show. So please click and subscribe and do all those things to sort of boost the show and share it with your friends.

And we'll see you next week. Thank you very much for watching. Thank You.