A body of academic systems work at FAST NUCES exploring concurrency, scheduling, and distributed computation from first principles.
SparkKernel — Multithreaded OS Simulation
A simulated operating-system kernel written in C++ using multithreading to model how a real kernel schedules and manages work.
- Custom process scheduler with First-Come-First-Serve, Round Robin, and Preemptive Priority strategies.
- Process & thread management following the classic five-state model.
- Kernel data structures — initializes processes/threads with their stacks and PCBs, tracking each process's state for statistics.
The five-state process model:
- NEW — being created, not yet executing.
- READY — ready to run, waiting to be scheduled on a CPU.
- RUNNING — currently executing on a CPU.
- WAITING — paused, waiting on an I/O request to complete.
- TERMINATED — completed.
MapMPI — Distributed MapReduce
A MapReduce framework implemented from scratch using MPI (Message Passing Interface):
- Distributed matrix multiplication and parallel Quick/Merge sorts across clustered nodes.
- Simulated Hadoop ecosystems to optimize grouping and indexing of large-scale academic-journal datasets by authorship metadata.
Tech Stack
C++ · Java · MPI · Hadoop · Multithreading