Artificial Intelligence thread

dripblackcoffee

New Member
Registered Member
What is the constraint here for scaling up? SMIC capacity, or the HBM? What is the realistic timeframe of gaining enough inference capacity to serve the world in a competitive fashion?
At the moment it seems like HBM, although its mostly everything chip related, probably 2035+ if we say everything goes perfectly (for china to serve everyone).
 

tphuang

General
Staff member
Super Moderator
VIP Professional
Registered Member
What is the constraint here for scaling up? SMIC capacity, or the HBM? What is the realistic timeframe of gaining enough inference capacity to serve the world in a competitive fashion?
That is a loaded question. In Kimi's case, it wants to be capital efficient, so if demand suddenly 2x or 3x, it's going to be compute constrained until it can sign new contracts.

For Alibaba for example, it has its chips/memory produced at Samsung. So there is actually no dependence on SMIC advanced node or CXMT HBM capacity. It's mostly a resource issue.
 

GiantCanofWater

Junior Member
Registered Member
What are the good or bads if they manage to ban Chinese models? Primarily for large s&p 500 companies since those are the only ones who can really afford the compute to run these giant models. Something like doing routine inspections on what their hosting so that even if someone touches up K3 and renames it “Trump Mind” they can still be caught? Will the ai companies actually be able to generate profit if their models are the only ones used? will the companies using their ai lose profit or just become less competitive due to the expense?
 
Top