Searching...
15 results for “hardware scaling”
Topic4 mentions across episodes
hardware
Concept2 mentions across episodes
hardware
...in time for the hardware. The depreciation cycle for the space power is more like somewhere between twenty five and forty years. So we have benefits there. I think if you think of on the networking side and you look at both enterprise and the hypersc
5:15 / 6:15
from transcript5:15 – 6:15
How quickly do you go from a single, replica of that model to five to 10 to a 100? And so that's the second that's the second pillar that is necessary for running these mission critical inference workloads. And what does it take to do that? It takes
33:00 / 34:00
from transcript33:00 – 34:00
...on, a set of hardware, is gonna get too much traffic that it cannot handle. And at that point, you need to horizontally scale it. And that's not an ML problem. That's not a PyTorch problem. That is, an infrastructure problem to ensure that you can ho
54:45 / 55:47
from transcript54:45 – 55:47
I don't think we're quite too back to mainframes in that it is still the case that people are running on scale out architectures across these pools. In other words, whether you have GPUs or TPUs, you're not necessarily saying, hey. That's my GPU supe
8:16 / 9:20
from transcript8:16 – 9:20
There's the cloud delivery layer with AWS, Azure. There's some of the new Neo clouds, and then some will be delivered on prem. And then above that is the foundational model layer. These are the big LLM companies like OpenAI, XAI, Google Gemini, Meta.
27:02 / 28:05
from transcript27:02 – 28:05
And also, no matter who wins above, we know we're gonna need tremendous amounts of AI infrastructure. And then third, we also know what the ecosystem's like. We know it's gonna happen on GPUs. We know it needs a lot of networking. And most of the die
27:43 / 28:49
from transcript27:43 – 28:49
...manufacturing, hardware enablers, the application layer may still generate significant long term winners, but in the near term, valuation compression suggests uncertainty around margin sustainability, competitive intensity. In contrast, the infrastru
2:52 / 4:03
from transcript2:52 – 4:03
As you think about long term systematic trends in aging populations and the requirements there are around serving that ecosystem, there are huge opportunities on that front. Similarly, in education, as you think about moves towards privatization, nee
6:01 / 7:06
from transcript6:01 – 7:06
private equity. Investment banking is all about deal velocity and how do I get this deal done, you put on the accelerator Right. Versus the private equity model, which is all about, look. My strong bias is not to do the deal. It's about deal quality,
9:44 / 10:51
from transcript9:44 – 10:51
you know, solve other other disciplines with. Right? Like, we solve grand unified field theory and physics goes away, and then we just go on to biology type thing. Right? So I do computer science. So infrastructure is the stuff used to build the apps
14:07 / 15:27
from transcript14:07 – 15:27
...hardware. Now we're doing sort of premium hardware, but it it'll become commoditized again, I assume. What kind of jobs are people throwing at the hardware now? And are there instances where they're overwhelming it or, you know, otherwise hitting a b
7:31 / 8:36
from transcript7:31 – 8:36
If it's an infrastructure essential service, you can be very confident in ten years' time. They've got more earnings and they're worth more than they are today, but the path to get there is not likely to be a straight line. That's Mike Dorril, the CE
0:05 / 1:10
from transcript0:05 – 1:10
So what this means is that the technology industry that's sort of willing this infrastructure into existence fundamentally needs to bring its own power to support that growth, which means massive investments, not just in data centers, but also in the
34:29 / 35:33
from transcript34:29 – 35:33
Have a podcast?
Run your podcast through Clypt
Get ranked clips, hooks, and ready-to-post copy from your own episodes. Free to try.