Unsupervised Learning with Jacob Effron
00:31:55 · Jacob Effron
[00:31:55] S1: Yeah. I mean, it seems that there's obviously a set of tools that folks are gonna use to build agents and build on open source. And it seems like you wanna have a bunch of those tools to be able to offer to Yep. To your end customers. And web search kind of being a a a key primitive for any any long running agent. We'll consume a decent amount of tokens on on just pulling information in. So Yeah. I guess I'd love to switch gears to just like the broader, you know, industry build out and and kind of the... You know, I I think there's lots of questions folks are asking about that. Figured one place to start that actually might just be interesting for our listeners is to what extent are, like, the data centers that you work with similar and, you know, like, there's a reference design and they all kind of, like, know, I don't know. To you, it doesn't really matter which one's which. And, like, what are some of the differences in them that actually impact the business of, you know, running an AI cloud on on top of them?
00:32:41 · Danila Shtan
[00:32:41] S0: To be frank with you, I have no idea what the walls of the facility are made of. So they're probably different, but that doesn't really help. But in all seriousness, what's within the walls is something that we fully control. We do not rent data centers when someone like have built the, like the network on their own or brought some GPU GPUs or servers, etcetera. If we don't own the facility and we rent it, for example, we rent, well, the building, electricity, but that's it. All the internal build out is controlled by us. It's like, it's done according to our standards. So from the software side of things, there is literally no difference which data center we run. Is it our own facility? Is it something that is built within a rented facility? Or for example, we've recently announced the so called asset light model when a third party brings capital in order to set up the facility and GPUs, and we run our platform on top of that. It's gonna be the same. These data centers are built according to our... Like fitted out according to our spec. Hardware is something that is familiar to us because like unknown hardware is quite hard to adopt and make sure that you provide proper SLAs. And that's it. So we don't compromise there. We never rent something from people and present it as our own cloud. If you launch a virtual machine in Nebius cloud, we can be sure that it runs on our stack top to bottom.
00:34:29 · Jacob Effron, Danila Shtan
[00:34:29] S1: Yeah. Makes makes sense. I mean, I guess there's a lot of dialogue around the the bottlenecks to the AI build out today. And I think people talk about a bunch of different ones, you know, whether it's power or labor to actually make, you know, data centers, cooling, networking, you know, HBM, GPUs, I guess, as as a as a overall. Like, how do you think about what the what the key bottlenecks are today? If you were to to kind of rank the... Across all of them, the the the few that feel most most poignant right now? [00:34:55] S0: I'm sorry to disappoint. There is one single bottleneck, and that's your ability to execute. Because, yes, all of those are issues. Like, there are GPU shortage in the market, components shortage in the Nowadays, it's CPU shortage in the market. Like NVMe shortage in the market. There are, like, not enough facilities, not enough electricity, not enough not enough qualified, like, workforce to help you build it out. Not enough, like, small components like network transceivers, for example. What else is not in abundance? Oh, oh, well, you know, like, those large scale... I'm not sure. I, like, I I know the proper English terminology, but, like, you know, those machines that actually convert electricity in place like transformators. [00:35:48] S1: Oh, transformers. Yeah.
00:35:48 · Jacob Effron, Danila Shtan
[00:35:48] S1: Oh, transformers. Yeah. [00:35:49] S0: Transformers. Yeah. Like, do you know the lead times on those these days? Oh, years. Right? So, like, everything is a problem. And I think the single bottleneck and at the same time, the single, like, mode that you can impose on that is, like, your ability to execute around all these problems and concerns, your ability to source, your ability to cut corners wherever appropriate, your ability to execute in general, your ability and flexibility to move stuff between facilities. Okay. We're gonna move this one from here to there, and then we're gonna move it back, kind of like, yes, we'll postpone this one, but this means we'll be in time with this one. This is important, etcetera, etcetera, etcetera. So I think that, like, this particular field, like an experienced project manager, is probably in highest demand I've ever seen throughout my career. [00:36:46] S1: Yeah. That's... I I guess the the the bottleneck is really good project managers, which Yeah. [00:36:50] S0: Exactly. Like, your your overall ability to, like, have this large context of, like, problems everywhere in mind and navigate that becomes a moat.
00:36:50 · Danila Shtan, Jacob Effron
[00:36:50] S0: Exactly. Like, your your overall ability to, like, have this large context of, like, problems everywhere in mind and navigate that becomes a moat. [00:37:03] S1: Yeah. It's interesting because you obviously have, you know, the hardware itself changes... Changing. You have a bunch of workloads changing. Obviously, we have much more long running agents. We have more post training going on, you know, people running open source models. How does that actually impact the way that, you know, the overall, like, I guess, the overall build out of the clusters themselves? To what extent is it like, hey, you know, a workload's a workload and, you know, we can we can run it?
00:37:27 · Danila Shtan, Jacob Effron
[00:37:27] S0: No. The workload's a workload. Like... Exactly like you said. It is slightly different, But in a sense, you need a high enough baseline to not be constrained by that. And then the rest is just your optimization problem. The way you pack things. The way you split things. Right? Like you don't need whole nodes for some of the workloads. So you need to collocate workloads. You need to have the right primitives. You need to expose those through the platform, etcetera, etcetera, etcetera. But in a sense, like differentiated workloads almost never like impose any changes on on on how you build on the lowest level. Yes, of course, like nowadays with like these new architectures, like what is it called? NVIDIA LPX, like the outcome of the grog deal. Right? This like adds to it. But in a sense, if you have experience running other generations, this is just another like 0.5 generation. Right? Yes, this will be a little bit novel and it's connected in a in a weird... Like, compared to the previous one's way. And it kinda is... Like, there are some restrictions when you can run and not run on that. But efficiency gains. And this is like just a, like a niche local addition to the platform to what to what you offer to the broader market. But in general, no. There is like way less custom, like custom solutions that you would expect from that. The workload is a workload. [00:39:17] S1: Why didn't the hyperscalers, like, move into this market faster or win this market, you know, more convincingly early on in in in this first chapter, at least?