Four HPC Platforms from Scratch. Here's What Actually Matters.
Lessons learned from building 4 complete HPC platforms from nothing — from CloudStack bare metal to SLURM for quant finance.
Thoughts on HPC, cloud infrastructure, and building platforms at scale
Lessons learned from building 4 complete HPC platforms from nothing — from CloudStack bare metal to SLURM for quant finance.
The growing gap between cloud-native and HPC skills, why AI is forcing convergence, and the career opportunity at the intersection.
What building HPC infrastructure at a quantitative hedge fund taught me about latency, jitter, and the difference between fast and predictably fast.
Building self-service HPC platforms at HPCFLOW and Millennium — what actually works, what users actually want, and why the ticket queue keeps coming back.
How HPCFLOW solved hardware-level network isolation for multi-tenant HPC — the project everyone said couldn't be done without destroying performance.
Demystifying InfiniBand for engineers who know TCP/IP but have never touched HPC networking — key concepts, RDMA, and when you actually need it.
A practitioner's view of how HPC infrastructure has changed since 2011 — what actually improved, what was hype, and why bare metal still matters.
Practical guide to getting into conference speaking as an engineer, with real examples from ISC, Supercomputing, and OpenInfra.
The founding engineer role isn't about being the first hire. It's about making the first technical decisions that everything else builds on. Lessons from building four platforms from zero.
I left university without a degree in 2011 and built a career in HPC. Here's what actually worked — and what I thought would work but didn't.