I applied through an employee referral. The process took 2 weeks.
First interview in which I mainly presented my thesis + general knowledge on LLMs (embeddings, tokens, transformers...).
Second interview in which I got a list of methods for parallel Distributed Deep Learning (like Data Parallelism, Model Parallelism...). For each method there was a short description (though it's suggested to be familiar with them beforehand), and the question was to calculate the memory complexity of each GPU according to the hyperparameters I got.
First interview, in which I mainly presented my thesis + general knowledge on LLMs (embeddings, tokens, transformers...). Second interview, in which I was given a list of methods for parallel Distributed Deep Learning (like: Data Parallelism, Model Parallelism...). For each method there was a short description (though it was suggested to be familiar with them beforehand), and the question was to calculate the memory complexity of each GPU according to the hyperparameters I was given.
The following metrics were computed from 1 interview experience for the Qualcomm AI Algorithms Engineer role.
Qualcomm's interview process for their AI Algorithms Engineer roles is extremely selective, failing the vast majority of engineers.
Candidates reported having very good feelings for Qualcomm's AI Algorithms Engineer interview process.