Shota M. | Air-Gapped Infrastructure Architect·Nov 3Eliminating Exit 137 OOM Crashes in 10M Batch Workloads: 9 Docker Flags, libjemalloc2 Injection…Executing 10,000,000 asynchronous LLM batch inferences or parsing 100GB+ binary files (ASPRS LAS, PCAPNG, ASAM MDF4, SEG-Y, BAM) on…
Vduvanenko·Jun 6Radix Selection AlgorithmThere is a closely related algorithm to sorting called Selection, which provides the k-th element from an unsorted array as if it was…
Ahmed Afzal·Apr 16From Bubbles to Full Utilization: A practicle Guide to pipeline Scheduling in Distributed Deep…Introduction
Shankar Ravi·Apr 4AI Fabrics: From System Lessons to Network IntelligenceThis article reflects my personal views and opinions only and does not represent the views of any past, present, or future employer.
Neethu u·Mar 1🚀 My Journey Into High-Performance Computing as a BeginnerSometimes, we choose a path not because it is easy, but because it challenges us.
Vimarshana Herath·Feb 7Architectural Analysis: Tiled Matrix Multiplication via CUDA C++Scaling dense linear algebra throughput requires a fundamental transition from the latency-optimized SISD model of the CPU to the…
DolphinDB·Dec 22, 2025C++ Speed Without C++ Pain: Inside a Microsecond-Level Low-Latency EngineIn ultra-latency-sensitive domains like quantitative trading and high-frequency data processing, every microsecond matters. What does it…
DolphinDB·Dec 18, 2025Making Scripts Fast with JIT CompilationProgramming languages typically fall into two categories: compiled and interpreted. Compiled languages like C/C++ translate source code…
OnComputing·Dec 1, 2025HPC Intro: Why HPC matters in 2025, and how it became a scientific pillarI think everyone remembers where they were when COVID-19 shut down the world. For me, the reality of the pandemic hit on March 11, 2020…
Syntal·Nov 27, 2025Java’s Vector API, ExplainedHow to squeeze GPU-like throughput from plain old CPUs — with safe, portable SIMD that actually ships.
