CUDA Fundamentals For Systems Interviews
A compact review of host/device code, grids, blocks, warps, memory hierarchy, synchronization, and CUDA compilation.
Read articleBlog
Technical notes on scientific computing, performance engineering, and the systems ideas behind research software.
GPU Performance Primitives: Tiling, Scan, Reduction, Roofline
Review notes for explaining GPU kernels through thread mapping, memory access, synchronization, and bottleneck hypotheses.
Read articleLock-Free MPMC Queues And Hazard Pointers
A walkthrough of Michael-Scott queues, CAS linearization points, memory ordering, ABA, and hazard-pointer reclamation.
Read articleC++20 jthread Pool: Futures, Wakeups, And Shutdown
The thread-pool layer around a queue: packaged tasks, futures, stop tokens, condition variables, graceful shutdown, and lifecycle locks.
Read articlePerformance Engineering Lessons From GF(2) Linear Algebra And CTMC
How bit-packing, M4RI-style table methods, layout experiments, and Fenwick-tree CTMC checks fit into one engineering loop.
Read articleAutograd For Systems Interviews
A systems-oriented refresher on reverse-mode autodiff, micrograd-style scalar graphs, PyTorch vocabulary, and GPU placement.
Read article