cat /etc/motd   Welcome to CMPUT 393 — Scalable Data-Intensive Analytics. University of Alberta · Department of Computing Science   Abstractions, algorithms, and infrastructure for scaling data science, ML, and data engineering across multiple machines.   * First offering: Winter 2027 * Status: under construction — the course is being built right now   uname -a cmput393.dev 0.0.2-prealpha #1 SMP PREEMPT_DYNAMIC many-machines x86_64   ls / syllabus/  labs/  project/  (nothing mounted yet — check back later)   man cmput393   CMPUT393(7)U of A Programmer's ManualCMPUT393(7)   NAME cmput393 — scalable data-intensive analytics   SYNOPSIS cmput393 [--winter-2027] --ai-first   LEARNING OUTCOMES Each outcome pairs what you will understand with what you will do to show it.   1.  You will understand the core concepts and trade-offs of data-intensive systems — replication, partitioning, storage and retrieval, consistency and fault tolerance, batch and stream processing — demonstrated by explaining a system's architecture and defending its design choices against alternatives, live, in discussion and debate.   2.  You will understand, as working intuition rather than formal analysis, the algorithms and data structures underneath data-intensive systems — such as consistent hashing, LSM-trees, and probabilistic structures like HyperLogLog — demonstrated by spotting them in the wild — in case studies, real codebases, and your own systems — and reasoning about which one fits a workload and why.   3.  You will understand what it takes to operate distributed infrastructure — provisioning, networking, security, and the ways real deployments fail — demonstrated by deploying multi-node systems on cloud infrastructure you administer yourself, and diagnosing and repairing the failures you meet along the way.   4.  You will understand how real systems scale — from one machine to cloud-native architectures: when to scale up, scale out, or not scale at all, and how partitioning, communication, and contention set the limits — demonstrated by analyzing case studies and real codebases, predicting and estimating how a design behaves under a given scenario before running it, and testing those predictions against measurement.   5.  You will understand how to evaluate a system honestly — designing workloads, choosing metrics, knowing how benchmarks mislead — and how to estimate what a system costs to run, demonstrated by building measurement harnesses, load-testing your own systems and your peers', pricing deployments at real cloud rates, and justifying design decisions on performance and cost together.   6.  You will understand the methodology of AI-first engineering — specifying intent, orchestrating the work, evaluating the result, and learning from the work as it happens — demonstrated by building your systems AI-first, deliberately and in the open: writing specifications before code, saying what you expect before you prompt and reading the result against it, accepting or rejecting generated work with stated reasons, and growing the class's community of practice by sharing what you discover.   SEE ALSO syllabus(5), labs(1), project(7) — pages not yet written   BUGS At scale, always.   AUTHOR Ildar Akhmetov <ildar.dev>   University of AlbertaWinter 2027CMPUT393(7)  
this host will also serve student deployments · uptime not guaranteed, that's the point