alt Hacker NewsThe unit economics seem pretty rough though. You're locking up 8xH100s for the compute of ~32B active parameters. I guess memory is the bottleneck but hard to see how the margins work on that.
The unit economics seem pretty rough though. You're locking up 8xH100s for the compute of ~32B active parameters. I guess memory is the bottleneck but hard to see how the margins work on that.
Yes, it only makes sense economically if you have batching over many users.