Gina Rosenthal talks Quantum, Server Refreshes, and Real-World AI Needs at HPE ProLiant Deep Dive

Well, it was interesting hearing about all of, uh, HP's, uh, ProLiant line, the new line. I think the most interesting thing I did not realize, because I've kind of been out of the server game for a minute about the quantum regulations that have already come out. Okay.

And I definitely didn't know. I, I figured they were looking at like 2030 to get these regulations established. Basically the idea being that from a security standpoint, that your server by 2030 had to be completely ready for a quantum um, environment.

And now they pulled back to 2027, which is two years from now, which from a server, um, manufacturing process is crazy. Usually server manufacturing takes seven to 10 years. They start like all the way back then.

So it sounds like HPE was new in what was coming and they were already working on the most important ones, which were up at the top, the important line items that were up at the top. And I, I think what, it'll probably be more than that. There's probably this two year timeframe ish, and I think the timeframe is more specifically people that are doing work with the federal government.

Yeah. Right. So they probably have seen this coming and they probably had already done the due diligence for, okay, what do we need to do by 2030 to get this ready?

And now what'll happen is I think the server vendors will make a bunch of money in the short term because it sounds like there's physical things and there's, there's got, this is just me guessing 'cause I haven't read the regulations, but there's got to be things that have, uh, software they'll stop, you know, kind of to hold 'em over for the next two years. But yeah, I think it, it is gonna be that race too. Okay, let's go install these new servers and how do we, it'll be, they probably already had this planned out when they wanted to do it for a big server refresh.

Mm-hmm. So now they'll just have to make sure the refresh lies within those parameters and Okay. Timewise everybody else thinks is boring.

I do not, I'm very interested in how they do the management. Um, and especially maybe if it's GreenLake, I don't know how they do the management and what's the underlying things. So how much, how are they actually doing the automation?

Is it Ansible, is it, you know, some other kind of custom automation stuff. So that would be interesting. But yeah, definitely looking forward to the AI discussion, you know, which I have a feeling is gonna be all Nvidia, which, you know, it's okay, but like, oh, I loved them showing the smaller footprint servers, the DL 1 31 40.

Right. 'cause if you're a, a small organization, you know, if you're a small business, that's perfect. You need a cluster just a little, you get a little rock, have everything in it.

Yeah. That just seems like so, such a good, um, prospect for smaller organizations are giving you something. At least if you, if you wanna try to do some development on premises and in use with the cloud, do it.

Um, but I just don't, you know, I wanna hear about, let's talk about real AI stuff that we're calling it AI now, but before we called it machine learning and Yeah. Deep learning, right? Like I don't need A GPU for some of that stuff, but it's still critical.

Yeah. You know, and, and those critical things. So let's talk about, you know, how can, what the servers are doing now and if it needs a GPU, okay, that's fine, but lots of stuff doesn't, and let's, we should start talking about that.