A school-cafeteria menu helper works on one intern laptop GPU and stalls when a class lines up. What should the team add before public lunch hours?
Select an answer to reveal the explanation.
Short Explanation
One intern laptop GPU is a prototype. A lunch line needs Triton or NIM in front for batching and concurrency. More print statements, a retrain on every tap, or an unnamed appliance will not scale the cafeteria.
Full Explanation
A script that works on one GPU is a prototype; a lunch line needs a serving layer such as Triton or NIM for batching and concurrency. Extra print statements and a larger notebook do not add that layer. Retraining on every tap, or buying an unnamed appliance, is the opposite of a scalable cafeteria path.