Preview
IB DP Computer Science — A4 Machine learning — Machine learning fundamentals | SL | A4.1.2
📋 What You Will Learn
- Describe the factors that determine ML hardware needs: processing, storage and scalability
- Distinguish the hardware demands of training from those of inference
- Describe the hardware spectrum from standard laptops to advanced infrastructure
- Describe GPUs, TPUs, ASICs, FPGAs, edge devices, cloud platforms and HPC centres and what each is best at
- Match a deployment scenario to an appropriate hardware configuration and justify the choice
🚚 Why This Matters You would not hire a fleet of articulated lorries to deliver one pizza, and you cannot deliver 10 000 parcels on a bicycle. ML hardware works the same way: a laptop happily trains a small model, while a large language model needs a warehouse of specialised chips — and choosing wrongly wastes money at one end or makes the project impossible at the other.
What Drives the Hardware Requirement
- Processing: ML — especially deep learning — is dominated by huge numbers of matrix