Interview practice

Deep Learning, OS, Embedded, C, Git, and NPU Q&A

This page is original study material generated from official documentation and durable course references. It is grouped for learning: fundamentals first, then debugging, deployment, and system-level tradeoffs.

How to use this page

Open one category at a time. First answer from memory, then expand the card. Search for keywords such as attention, deadlock, DMA, linker, or volatile. The separate analog CIM study adds Mythic architecture, evidence review, and interactive MCQs.

Source Map

Deep learning: PyTorch neural networks, Dive into Deep Learning, CS231n notes, Hugging Face Transformers.
Operating systems: Linux man-pages fork(2), Linux memory management, Linux MM concepts, OSTEP.
Embedded systems: FreeRTOS queues, FreeRTOS mutexes, Arm linker scripts, Arm volatile MMIO note.
NPU platform engineering: SoC software architecture, framework, compiler, ISA, and tooling skills, compiler practice.
Analog CIM: Mythic architecture, accuracy evidence, hardware/software co-design, Tenstorrent reuse decision, role study, MCQ lab.