11–20 of 390 posts
The bubble pops when Apple releases an iPhone that runs a good enough for most things LLM locally. At that point cloud hardware investments will plateau (unless some new GPU melting use case comes out). Investors will move from nvidia, AMD into Apple.
The bubble pops when Apple releases an iPhone that runs a good enough for most things LLM locally. At that point cloud hardware investments will plateau (unless some new GPU melting use case comes out). Investors will move from nvidia, AMD into Apple.
I wonder what we will do with all that compute if no one is going to use it...even if we achieve AGI it won't run on this generation of hardware. Are we just gonna brute force architectures or wtf is going on?
Cheap cloud gaming. Those GPU's will finally push pixels again :)
I wonder what we will do with all that compute if no one is going to use it...even if we achieve AGI it won't run on this generation of hardware. Are we just gonna brute force architectures or wtf is going on?
Cheap cloud gaming. Those GPU's will finally push pixels again :)
> Those GPU's will finally push pixels again :)
Have they ever? I wonder if, say, an H100 even supports graphics APIs.
The bubble pops when Apple releases an iPhone that runs a good enough for most things LLM locally. At that point cloud hardware investments will plateau (unless some new GPU melting use case comes out). Investors will move from nvidia, AMD into Apple.
What's the benefit to running LLMs locally? Data is already remote, LLM inferencing isn't particularly constrained by Internet latency. So you get worse models, performance, and battery life. Local compute on a power constrained mobile device is required for applications that require low latency or significant data throughput and LLM inferencing is neither.
At work:
That I don't rent $30,000 a month of PTUs from Microsoft. That I can put more restricted data classifications into it.
> LLM inferencing isn't particularly constrained by Internet latency
But user experience is
AMD just doesn’t know how to compete with Nvidia. The best it can do is charge 10% less and release GPUs about the same level of underwhelming performance as Nvidia.
Maybe they wouldn’t need to sell equity if they made better faster cheaper products than Nvidia.
But Lisa Su can’t bring herself to compete. Which is very strange because she brutally competed with and destroyed Intel.
The bubble pops when Apple releases an iPhone that runs a good enough for most things LLM locally. At that point cloud hardware investments will plateau (unless some new GPU melting use case comes out). Investors will move from nvidia, AMD into Apple.
What's the benefit to running LLMs locally? Data is already remote, LLM inferencing isn't particularly constrained by Internet latency. So you get worse models, performance, and battery life. Local compute on a power constrained mobile device is required for applications that require low latency or significant data throughput and LLM inferencing is neither.
The general advantage is that you know that you're not leaking information, because there's nowhere to leak it to. You know the exact input, because you provided it. You also get the benefit of being able to have on device encryption, the data is no good in the datacenter if it's encrypted.
The bubble pops when Apple releases an iPhone that runs a good enough for most things LLM locally. At that point cloud hardware investments will plateau (unless some new GPU melting use case comes out). Investors will move from nvidia, AMD into Apple.
Earlier quoted context omitted.
What's the benefit to running LLMs locally? Data is already remote, LLM inferencing isn't particularly constrained by Internet latency. So you get worse models, performance, and battery life. Local compute on a power constrained mobile device is required for applications that require low latency or significant data throughput and LLM inferencing is neither.
> What's the benefit to running LLMs locally? At work: That I don't rent $30,000 a month of PTUs from Microsoft. That I can put more restricted data classifications into it. > LLM inferencing isn't particularly constrained by Internet latency But user experience is