3 docs tagged with "tradeoffs"

| Store | Local / Hosted | Setup effort | Metadata filtering | Scale ceiling | Cost model | Pick it when |

Most failed GPU ports do not fail because the kernel was slow. They fail because the workload was never shaped like something a GPU accelerates, and the port made that visible only after weeks of work. The kernel itself often does run twenty times faster than the CPU loop it replaced — and the program gets slower anyway, because the time now goes into transfers, synchronization, and the 60% of the runtime that was never offloaded at all.

A framework earns its keep when it removes real complexity. For the simplest LLM use case — one