For two decades the word "supercomputer" meant one thing: a Cray. Then commodity microprocessors caught up on price/performance, the killer micros arrived, and the custom vector machine died. This is how vector processing worked, why it lost, and what crawled out of the wreckage to become modern HPC.
Hpc
-
The Cray Era and the Death of the Supercomputer -
Fortran Is Not Dead Fortran still runs the world's most important numerical simulations. A deep technical look at why the language endures, how modern Fortran 2008/2018/2023 differs from FORTRAN 77, the parallelism story, the compiler landscape in 2026, and the honest costs of working in the ecosystem.
-
Building Modern Toolchains on Enterprise Linux: Python, Compilers, and Utilities on SLES 15 and Friends How to give users modern tools — recent Python, compilers, and CLI utilities — on deliberately-frozen enterprise distros like SLES 15 and RHEL, comparing Spack, EasyBuild, Nix, conda, and containers without breaking the stable base.
-
Environment Modules with Lmod: How `module load` Actually Works How `module load` actually works — Lmod as a thin Lua layer that rewrites your shell environment — and how to build a hierarchical software tree that serves hundreds of HPC users with conflicting toolchains and no filesystem collisions.
-
MPI Programming Essentials: Collectives, Non-Blocking, and the Traps Nobody Warns You About A working engineer's guide to MPI: the mental model, point-to-point and collective communication, non-blocking patterns, derived datatypes, one-sided RMA, threading levels, MPI-IO, performance traps, and integration with SLURM.
-
NCCL Deep Dive: Multi-GPU Collectives, Ring vs Tree, and Debugging Distributed Training How NVIDIA's NCCL moves gradients across multi-GPU and multi-node training — ring versus tree collectives, NVLink and InfiniBand topology awareness, and how to debug the silent hangs and bandwidth cliffs that take training down.
-
OpenMP: Threading Without Pthreads (And Without the Nightmares) OpenMP as the sane way to parallelize C, C++, and Fortran — pragmas instead of pthread plumbing for parallel loops, tasks, and GPU offload — plus the data-sharing and scheduling details that decide whether you get speedup or bugs.
-
Parallel Filesystems Compared: Lustre vs BeeGFS vs GPFS vs Weka When NFS stops scaling you need a parallel filesystem — a comparison of Lustre, BeeGFS, GPFS, and Weka covering how each distributes data and metadata, their performance profiles, and the operational burden each one brings.
-
RDMA and InfiniBand From the Ground Up: Why It's Fast and How to Diagnose It Why RDMA and InfiniBand deliver 400 Gbps at single-microsecond latency with the CPU idle — how kernel-bypass and the verbs model work, how RoCE differs, and how to diagnose a fabric that isn't hitting its datasheet.
-
Spack and EasyBuild: Reproducible HPC Software Stacks Without Losing Your Sanity Reproducible HPC software stacks with Spack and EasyBuild — treating every build as a versioned recipe that resolves dependencies and generates coherent module trees, so your cluster's software doesn't become an unrepeatable mess.
-
SLURM: The HPC Job Scheduler — From Basics to Advanced Administration A comprehensive deep-dive into SLURM — history, architecture, user commands, advanced workflows, cluster administration, and tracking job efficiency.