The Six Five team discusses HPE Discover 2023.
If you are interested in watching the full episode you can check it out 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:
Daniel Newman: Let’s talk HPE Discover 2023.
Patrick Moorhead: Yeah, so you and I spent three days there. I rolled in Monday morning and I guess we left Wednesday afternoon. But no, these big tent events for all manufacturers are a time that I think it’s important to soak in the strategy, compare to next year, weigh in on the big announcements they had. Now I want to focus in on two. So HPE was the first to say it was going as a service for all on-prem data centers, server storage and networking. And then they’re adding software based solutions to the HPE GreenLake portfolio on a routine basis. But one thing that I’ve been following pretty closely, and you know I love the hybrid multi-Cloud, is over time the context that the public Cloud is 19 years old. At what point are enterprises going to more seriously consider for each workload having an on-prem as a service or edge as a service as part of its hybrid multi-Cloud strategy?
We already know that between 75 and 90% of all data and investment is still on-prem. But at what point does the hybrid Cloud really kick in? And one of the key players in this is a company called Equinix. And Equinix owns property crate. Basically it’s their data center as a service. So whether you’re HPE, Dell, Citibank, Wells Fargo, if you don’t want to operate a data center with the cooling capabilities, the electrical capabilities, all the real estate and the investment, you go to companies like Equinix and Digital Realty. So the big announcement here was that HPE and Equinix launched private Cloud portfolio at Equinix. So HPE is actually pre-deployment hardware at seven key metro areas around the world in Sydney, Singapore, Frankfurt, Washington, DC, Toronto, and Silicon Valley. I don’t know, does that mean Oakland? I have no idea, but you get the idea.
And if you think about this and why I think this is so provocative, it removes yet one more objection. Now, HPE had always aligned with Equinix, but it was spotty. And essentially what you can do just like you do with an AWS or a GCP or an Azure, if you want to go all in and have quote, unquote, “availability zones” around the world at those seven places that I talked about, you can do this. And here’s the other thing, instead of waiting one, two weeks a month for that hardware to show up, HPE has pre-bought and pre-installed hardware in there for you to go in there. And I was hearing timeframes of a day, which is just so much better. So we’re getting closer to those capabilities of the… Not all of them. You don’t have kind of unlimited scalability like you do at an AWS, but imagine though, Daniel, you add a multi-tenant system and a credit card swiper, you could be a developer and you could sign up for HPE GreenLake, do all of your testing and all of your… I mean, within seconds or minutes.
So pretty excited about this. And this is kind of the long track of the hybrid Cloud and the multi-Cloud coming together. That was announcement one. And the second one, and I won’t go too deep, leave you oxygen in the room was, HP announced an LLM as a service on-prem using its supercomputers. The company really didn’t lean in too much that, hey, this was going on a Cray supercomputer that has this incredible interconnects that connect different clusters because I think they want to have a little flexibility in what they do in the future there. But it seemed to me a very well thought out offering so far for very highly technical customers. And I think we already see HPE ingrained into governments, into three letter acronym security companies, and these are the kind kind of companies that I get the sense of because HP didn’t roll out any customers.
These are the type of customers I think will be most attracted because it’s on Cray. And whenever you’re programming for a specific environment, you have to program to, I think it’s called CDE with Cray, that enables you to bop back and forth between multiple clusters. It’s single tenant by the way. Overall it’s multi-tenant, but what you’re doing is essentially having one, two, 10, 100 different clusters that you can sign up for. I have a lot of questions because there’s a lot of things that weren’t disclosed but don’t know pricing, don’t know the basis of the pricing. Is it by the cluster? Is it by the cluster by GPU? A lot of open-ended questions right now.
I also don’t have a sense of what it takes to program to CDE versus let’s say something from AWS. And I also don’t know when GA is going to be over. This is not a paper launch, it’s also not GA like IBM’s supercomputer training as a service, but it’s a good start. This is something they’re offering that Dell doesn’t offer and I don’t think Lenovo offers it. So it seems like a differentiator to me. It also seems very much in line strategically with what the company wants to do and their continued progress with on-prem IAS.
Daniel Newman: Yeah, I think you hit on a couple things. I mean, one is thematically speaking, I thought the best moment was the moment when Antonio Neri on stage called it hybrid Cloud by accident.
Patrick Moorhead: Yeah.
Daniel Newman: Great quotes. And really talking about how the company was declaratively ahead of the curve on one, landing and hybrid Cloud. Two, the need to build this consumption based service. on-prem because I’ll say it now, I’ll say it again, I’ll continue to reiterate it. The operating model of Cloud is the desire. The actual location of the infrastructure, the desire varies. It varies by company need, but the ability to function and operate that way and HPE does and continues to have a substantial lead in terms of its service distribution. The crawl chart of services for HPE GreenLake is still far ahead of its competition.
I’ve said before, and I’ll say again, I don’t believe that lead will necessarily be sustainable forever. You’ve got very capable, competent competition and you also have the forces of the hyperscale Cloud that sees prem as a massive opportunity, whether it’s Andy Jassy saying 90% of workloads are still on-prem or it’s Antonio Neri saying 70% of enterprise workloads are still on-prem, or it was our guest on the pod yesterday from Citi that suggested that at some point, Cloud workloads will exceed prem workloads.
That period of time in that transition’s going to be long, and we are going to see services that are going to basically bring the two things together. So there was a significant sort of opportunity for HPE to really reiterate and lean into the fact that it’s been on this, it’s been ahead of this, and it continues to be. And I mean, it deepened its AWS partnership, it announced that from the stage.
Patrick Moorhead: Yes.
Daniel Newman: It leaned into a VMware partnership, which continues to have a lot to do with the fact that the world is hybrid. So the company wasn’t running away from hybrid by any means and wasn’t running away from workloads being in the public Cloud. So that that’s mostly just validating that we’re right, which is important to me to do at least once per topic, per podcast.
Patrick Moorhead: Per hour.
Daniel Newman: Yeah, I mean, the second thing by the way is-
Patrick Moorhead: Hey, Dan, though. Hey, was that interaction between Matt Wood at AWS and Antonio a little bit awkward when you kind of listened to the words that Matt used, and I hadn’t heard had AWS use this phraseology before, but basically the way that I internalized it was Matt said, “Hey, everything’s going to the public Cloud, and this is just like an intermediary step to help our customers get them there.” And it’s like, I haven’t heard anybody but Andy Jassy use that phraseology, but man, like you’re on HPE’s stage and you’re saying this. I just thought was very provocative. I don’t know if anybody else caught it, but I thought it was a little awkward.
Daniel Newman: I didn’t catch it, but now that you say it, there’s a little bit of room for reflection. I mean, look, that’s the bet. That’s AWS’s bet, but it’s sort of their bet because Outpost, they still know they’re going to have to put their infrastructure, but I think from a control plane standpoint, they’ve always sort of had the approach like, we want to own it all. And so I don’t necessarily see that course changing and I see the partnership as valuable to HPE. Just a quick touch on the LLM thing. It seems early to me, it seems very positive and promising that the company’s going to be able to do this, and it does sort of validate whether it’s been Bluedata, whether it’s been Cray, this company is not new to AI and I think some people want to make the assessment that HPE has just sort of jumped into AI out of nowhere.
I mean, they’ve actually been building architecture and like I said, they’re data services for GreenLake are far ahead of any of the other prem Cloud. It doesn’t even make sense, but the consumption-prem offerings that are out there, and it’s been going down the path of AI, whether that’s been operational AI for ML and AI ops, all the way to the super computing on demand. So I think there’s story there. And then of course they partnered with Luminous, ominous. It’s ominous that they worked with Luminous, but they partnered with a different LLM company, which is kind of interesting too. But a lot of what that LLM focuses in on are things that are important to enterprises. So some of the-
Patrick Moorhead: Have you ever heard of that company before?
Daniel Newman: I was not familiar Pat, but when you look deeper into it, it’s got a rich heritage. It’s been doing this for some time and it is focused a lot on transparency, which is interesting because this enterprise transparency issue is, where was your data? Where were these sort of generative assets constructed from? Where did the… It’s got a lot of that provenance built into the software to function a little bit differently than some of the other open source and popular LLMs.
Patrick Moorhead: I would’ve loved a superfluous Hugging Face announcement alongside a company that I had never heard of before.
Daniel Newman: Well, you got to do something with Hugging Face. I think it’s a requirement right now.
Patrick Moorhead: No, I know, totally. That’s what I mean. Even if it’s a superfluous fly by like, “Hey, and we support Hugging Face models too.”
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.
