Fundamentals of Large Language Models →mediumMultiple SelectObjective-mapped

1Z0-1127-25 Fundamentals of Large Language Models Practice Question

Which TWO techniques are commonly used to reduce the memory footprint of LLM inference?

⚠ Common exam trap

Oracle often tests the distinction between training and inference techniques, so candidates mistakenly apply gradient checkpointing (a training memory saver) to inference, or confuse batch size scaling with memory reduction.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

✓

Quantization

Quantization reduces the memory footprint by lowering the precision of model weights and activations from FP32 to lower bit-widths like INT8 or FP16, which directly decreases the memory required to store and compute with the model. KV cache optimization reduces memory usage by efficiently managing the key-value cache during autoregressive decoding, often through techniques like shared memory, pruning, or compression, which is critical for long-context inference.

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

✓
Quantization
Why this is correct
Reduces memory by using lower precision weights.
✗
Increasing batch size
Why it's wrong here
Larger batch size increases memory usage.
✓
KV cache optimization
Why this is correct
Reduces memory for storing key-value tensors during autoregressive generation.
✗
Gradient checkpointing
Why it's wrong here
Gradient checkpointing trades compute for memory during training, not inference.
✗
Using full precision (FP32)
Why it's wrong here
Full precision uses more memory than lower precision.

Go deeper

1Z0-1127-25 Fundamentals of Large Language Models Practice Question

Why each option matters

Option-by-option breakdown

Related to this question