No that makes no sense. They only need a tiny device to act as a proxy. It makes no sense to colocate the expensive GPUs
replyYeah, the usual approach here is to offer a free VPN service, and then piggyback traffic over the user’s internet provider. Much, much more scalable and cheaper than offering to put GPUs at a user’s home.
replyIt makes sense when someone else is paying for the power. The AI capabilities of mobile cpus is for pushing cloud based AI to your device. You pay the power bill while retaining the same lack of privacy as cloud hosted AI.
replyI'm pretty sure the company that owns the compute is the one paying for the power. Who would agree to have an ugly noisy cube in their property that they have to pay maintenance for? What would they get out of that deal?
reply>also why hyperscalers are paying folks to host gpu clusters in their homes
replySource? There's definitely shady people paying people to host proxies, but hyperscalers doing it would be surprising.
My understanding of it so far is that it is at most, 50 or 100 houses as a test... It's a startup that wants to make it wide scale but hasn't got there yet.
replyhttps://www.google.com/search?client=firefox-b-d&q=span+resi...
Calling them a "hyperscaler" is a stretch. By all accounts it's some renewables startup that's trying to pivot into AI because obviously AI = $$$, and calling themselves a "hyperscaler" in the process. Calling them a "hyperscaler" makes as much sense as some guy who runs a homelab in his basement as a "datacenter".
replyhttps://www.span.io/blog/span-announces-xfra-a-distributed-d...
My thoughts exactly. In the current environment you could slap "AI" on a potato and get $$$ for it.
reply