Distributed Systems

• GPUDirect Storage in 2025: Optimizing the End-to-End Data Path · 2025-09-16

  How modern systems move data from NVMe and object storage into GPU kernels with minimal CPU overhead and maximal throughput.

• From MapReduce to Spark: The Arc of Data-Parallel Systems · 2025-05-19

  MapReduce taught fault-tolerant batch at scale; Spark generalized it with resilient distributed datasets (RDDs) and DAG scheduling.

• Exactly-Once in Streaming: What It Means and How Systems Achieve It · 2025-01-22

  Disentangle marketing from mechanisms: idempotence, transactions, and state snapshots behind ‘exactly-once’.