For single user use, memory bandwidth is usually the bottleneck. To saturate the compute, you usually need batch inference, I.e. multiple users chatting at the same time or paralell processing of lots of documents. GPUs are much faster for this.
Memory bandwidth of 3090 is 936 GB/, 4090 about the same. Strix Halo should be around 256 GB/, a bit less than Macs with 128GB ram and about 1/4 of 3090/4090. Should be fast enough to be usable. 5090 has 1,792 GB/, so quite a bit faster.
Would it be possible to create some decentralized version of operating system? EG a version of linux married with nodes, with the whole filesystem hosted on Autonomi?
I think where we are going humans wonāt use an OS any more 
For AI you might want to wait for this : " NVIDIA Project DIGITS With New GB10 Superchip Debuts as Worldās Smallest AI Supercomputer Capable of Running 200B-Parameter Models". Price ~3000$, should be available in May.
This has also good performance / $ : NVIDIAĀ®Jetson AGX Orinā¢ 64GB Developer Kit
Itās a dev kit, but who cares, you can run quite many LLM on it.
yeah was looking at these nvidia jetsons the other day.
Jetson Orin Nano Super
was wondering if they could run autonomi nodes
is possible to connect android phone with trezor and uniswap?
Im curious why no one is discussing current prices. Is there another thread for this?
It is over in this thread:
Also check discord, there is a channel for tolken talk there.
And that thread is currently totally off-topic to be honest 
Itāll get back on topic soon enough
4 posts were merged into an existing topic: ANT Token - Price & Trading topic
Closing this topic, please continue discussions in the ANT trading topic here