Train your team

AIC-310 · GPU & AI Compute

Storage & Data Pipelines for AI

Stop starving your GPUs: parallel filesystems, GPUDirect Storage and pipelines built for sustained throughput.

Practitioner 3 days in person6 half-days online Max 14 in person

Who this course is for

Engineers responsible for keeping training jobs fed — the storage systems, formats and data pipelines that decide whether GPUs wait on data or not.

Prerequisites

Linux administrationBasic storage concepts (RAID, filesystems)AIC-200 or AIC-230 helpful

Course outline

Day 1 — Storage systems for AI

  • RAID levels and what they actually protect
  • NVMe and NVMe-oF
  • Parallel filesystems: Lustre, GPFS, BeeGFS
  • Object storage: MinIO, Ceph; tiering and archival
  • I/O optimisation for sustained (not burst) training throughput

Day 2 — High-performance data movement

  • GPUDirect RDMA and GPUDirect Storage
  • Zero-copy movement: GPU ↔ NIC ↔ storage
  • Parquet and Arrow formats for ML data
  • Dataset caching strategies: node-level and smart caching
  • Prefetching and pipelining into GPU pipelines

Day 3 — Data pipelines for training

  • DataLoader optimisation: workers, prefetching
  • Distributed data loading and sharding
  • WebDataset format; NVIDIA DALI for GPU-accelerated loading
  • Data augmentation on GPU
  • Memory-mapped datasets; checkpoint-resume for large datasets
  • Capstone workshop

Hands-on labs

Labs follow the academy model — 35% principles, 20% guided investigation, 45% engineering studio. Every claim you make in a lab is backed by a trace, a counter or a measurement you captured yourself. How we teach

  1. Lab: benchmark storage I/O against a training-style read pattern; identify where sustained throughput collapses
  2. Lab: build a Parquet/Arrow dataset and measure load throughput vs naive formats
  3. Lab: optimise a PyTorch DataLoader (workers, prefetch, pin_memory) until the GPU stops waiting; then replace it with DALI
  4. Lab: implement a node-level dataset cache and measure its effect on epoch time

Capstone project

Diagnose and fix a starved training pipeline: given a job whose GPUs idle on data, you trace the bottleneck through storage, format and loader layers, then deliver a pipeline that sustains a target throughput with measurements at each layer.

What you leave with

  • Storage selection and benchmarking skills for AI workloads
  • Working GPUDirect/DALI/Parquet pipeline experience
  • A layered methodology for finding data bottlenecks
  • Caching and sharding patterns you can deploy immediately

How it runs

Every course follows the same model: 35% principles, 20% guided investigation, 45% engineering studio. You leave with working code, raw measurements and an evidence-based report — not a certificate of attendance. Read the methodology or see a full sample lesson.

Material is adapted to your kernel version, hardware and workload before a private delivery. For public cohorts, the environment is provided and configured.

Questions

Who is this course for?

Engineers responsible for keeping training jobs fed — the storage systems, formats and data pipelines that decide whether GPUs wait on data or not. It sits at practitioner level within the GPU & AI Compute track.

What do I need to know already?

Specific prerequisites for this course: Linux administration; Basic storage concepts (RAID, filesystems); AIC-200 or AIC-230 helpful. We confirm levels before the cohort starts and adapt if a group is stronger or weaker than expected.

Can this run privately for my team?

Yes. Any course runs on-site at your offices anywhere, or live online for a distributed team, with labs adapted to your hardware and codebase.

What is the difference between in-person and online?

In person is 3 full days with hardware on your desk, capped at 14. Online is 6 half-day sessions across about two weeks so you can keep working, capped at 20, with remote lab access.

Do you invoice companies?

Yes. Purchase orders are accepted and invoicing is available in USD, EUR, GBP, SAR and CAD.

Upcoming dates

DatesWhereSeatsEarly birdRegular
22 Nov – 24 Nov 20263 full days RiyadhIn person · KAFD Conference Centre 3 of 14 SAR 7,090until 23 OctSAR 7,880
22 Nov – 24 Nov 20263 full days Kuwait CityIn person · Al Hamra Tower 8 of 14 KWD 580until 23 OctKWD 650
29 Nov – 1 Dec 20263 full days MuscatIn person · Knowledge Oasis Muscat 3 of 14 OMR 730until 30 OctOMR 810
6 Dec – 13 Dec 20266 half-days Gulf bandLive online · 09:00–13:00 GMT+3 3 of 20 US$1,350until 6 NovUS$1,500
7 Dec – 9 Dec 20263 full days OttawaIn person · Kanata North Tech Park 8 of 14 CAD 2,570until 7 NovCAD 2,860
7 Dec – 9 Dec 20263 full days TorontoIn person · MaRS Discovery District 3 of 14 CAD 2,570until 7 NovCAD 2,860
7 Dec – 14 Dec 20266 half-days Europe bandLive online · 09:00–13:00 CET 8 of 20 US$1,350until 7 NovUS$1,500
14 Dec – 16 Dec 20263 full days LondonIn person · Shoreditch Works 8 of 14 GBP 1,480until 14 NovGBP 1,640
14 Dec – 21 Dec 20266 half-days Americas bandLive online · 13:00–17:00 ET 13 of 20 US$1,350until 14 NovUS$1,500
21 Dec – 23 Dec 20263 full days BerlinIn person · Factory Görlitzer Park 3 of 14 EUR 1,740until 21 NovEUR 1,930

Dates shown for the next few months. If nothing fits, tell us where and when — cohorts are added on demand, and private delivery can be scheduled any week.

More in GPU & AI Compute

AIC-1003 days Foundations for AI Compute The architecture, operating system and networking groundwork every GPU systems engineer is assumed to have and often does not. Foundation Practitioner-taught SAR 6,750Next 25 Oct AIC-1103 days GPU Architecture, Memory & Interconnects How the hardware constrains your workload: SIMT execution, the memory hierarchy and the fabric between GPUs. FoundationNCA-AIIO Practitioner-taught SAR 6,750Next 11 Oct AIC-2004 days Linux for GPU Systems What the kernel is doing underneath your training job, and how to tune it. The layer almost nobody teaches. Practitioner Practitioner-taught SAR 10,500Next 15 Nov AIC-2104 days RDMA & AI Cluster Networking Build and debug the fabric distributed training runs on, from queue pairs up to a tuned NCCL all-reduce. PractitionerNCP-AIN Practitioner-taught SAR 10,500Next 1 Nov AIC-2205 days CUDA & HIP Programming Write, profile and optimise GPU kernels on both vendors, including the CUDA-to-HIP porting path. Practitioner Practitioner-taught SAR 13,120Next 11 Oct AIC-2304 days Distributed Training with PyTorch From a single-GPU training loop to sharded multi-node training that survives a node failure. Practitioner Practitioner-taught SAR 10,500Next 15 Nov AIC-3004 days Containers, Kubernetes & GPU Schedulers Run a shared GPU cluster multiple teams can actually use: partitioning, scheduling, quotas and isolation. PractitionerNCP-AII Practitioner-taught SAR 10,500Next 18 Oct AIC-3203 days MLOps & Inference Serving Get models off a laptop and onto a GPU endpoint that scales, with the pipeline machinery around them. PractitionerNCP-AIO Practitioner-taught SAR 7,880Next 1 Nov AIC-3302 days AI Infrastructure Security & Observability Harden a multi-tenant GPU platform and see what it is doing before users report a problem. Advanced Practitioner-taught SAR 6,000Next 18 Oct AIC-4003 days Large-Scale Training & Datacenter Architecture The thousand-GPU conversation: 3D parallelism, reference architectures, TCO and where the hardware is heading. Advanced Practitioner-taught SAR 9,000Next 8 Nov