These are living systems-engineering courses rather than polished textbooks. I use them to build connected mental models, work through debugging and measurement, and turn questions or failed assumptions into better explanations and experiments.

In development

AI Infrastructure Firmware Engineering Specialization

A larger, connected curriculum is taking shape around the firmware and platform systems behind AI infrastructure: OpenBMC and UEFI, PCIe, NIC firmware and RDMA, hardware trust, rack onboarding, continuous host assurance, fleet updates, observability, and serviceability.

The specialization will reuse and connect the focused courses below instead of duplicating them. Its public course board tracks curriculum design, lessons, labs, expert review, and interviews while the material is being developed.

Active

NIC Firmware Engineering

A systems-level course on how modern NICs move packets between wire, device, PCIe, memory, kernel, and application software.

The emphasis is on the boundary between hardware and firmware/software: DMA engines, queues, descriptor ownership, interrupts, PCIe transactions, cache behavior, multiqueue scaling, driver architecture, and debugging.

Performance & Cache Optimization

A systems-first course organized around tracing memory accesses end-to-end: virtual address generation, TLBs and page walks, cache lines and associativity, DRAM, coherence, prefetching, memory ordering, atomics, and performance experiments.

The goal is to connect CPU architecture to practical optimization rather than treating caches, virtual memory, concurrency, and profiling as isolated topics.

OpenBMC & UEFI Platform Firmware

A systems-first course on server management and boot firmware: BMC hardware, OpenBMC Linux, Yocto, D-Bus, Redfish, IPMI/KCS, MCTP/PLDM, UEFI, power sequencing, firmware update, security, debugging, and fleet-scale reliability.

The course uses external material such as Arm’s RD-V3 OpenBMC/UEFI learning path as a practical companion while keeping this repository as the canonical personalized textbook.

How the coursework evolves

Each lesson can carry browser-local review comments beside section headings. Questions, unclear explanations, correction notes, diagram requests, and lab ideas can be collected into a Markdown report and fed back into the next revision. See how the review system works →