Get started
Start with Mathematical Foundations
Vectors & Geometry
Back to modulesModule 15 · 0/37
Module 15
AI Systems & MLOps
0/377273
Submodule 01Data Pipelines, Monitoring & Reliability0/20
- 1Build a Simple ETL Pipeline (MLOps)Med.
- 2Analyze Canary Deployment Health for Model RolloutconceptMed.
- 3SLA Compliance Metrics for Model ServiceconceptEasy
- 4Feature Drift Detection using Population Stability IndexconceptMed.
- 5Prediction Distribution MonitoringconceptMed.
- 6Batch Prediction Health MetricsconceptEasy
- 7Model Inference Statistics for MonitoringconceptEasy
- 8Cold Start Latency Budget BreakdownconceptMed.
- 9End-to-End Latency DecompositionconceptMed.
- 10P50/P95/P99 Latency PercentilesconceptEasy
- 11Runtime Gym Space ValidationconceptEasy
- 12Frame-Aware Corrupt for Drift SimulationconceptMed.
- 13Cache-Aware Request Routing Across ReplicasconceptMed.
- 14Analyzing Memory Fragmentation in LLM ServingconceptEasy
- 15Autoscaling Replica Simulator with SLA TrackingconceptMed.
- 16Request Batching for InferenceconceptMed.
- 17ML Pipeline DAG Scheduler with Critical Path AnalysisconceptHard
- 18Critical Path Method for SchedulingconceptMed.
- 19Resource Dilation Factor for SchedulingconceptMed.
- 20Iterative Repair for SchedulingconceptMed.
Submodule 02Infrastructure, Parallelism & Hardware Efficiency0/17
- 1Estimate Minimum GPU Count for Model DeploymentconceptEasy
- 2Multi-GPU Communication Overhead (NVLink vs InfiniBand)conceptMed.
- 3Tensor Parallelism All-Reduce Communication CostconceptMed.
- 4Context Parallelism with Ring Attention for Video ModelsconceptHard
- 5Designing 4D Parallelism for Large Model TrainingconceptHard
- 6Multi-Instance GPU (MIG) Resource AllocationconceptMed.
- 7GPU Ops:Byte Ratio Calculation from Spec SheetconceptMed.
- 8Roofline Model Analysis for GPU OperationsconceptMed.
- 9Copy-on-Write Memory Sharing for LLM SamplingconceptMed.
- 10Kernel Fusion Memory Savings CalculatorconceptMed.
- 11Cython vs Python Performance for Array OperationsconceptMed.
- 12IPC Methods Performance ComparisonconceptMed.
- 13Lock-Free Synchronization with Shared FlagsconceptMed.
- 14Shared Memory Circular Buffer for RLconceptMed.
- 15Shared Memory IPC for RL EnvironmentsconceptMed.
- 16Zero-Copy Batch Data Loading from Shared MemoryconceptMed.
- 17Zero-Copy C Environment ImplementationconceptMed.