Gemma 2 9B IT: how much VRAM does it need?

Gemma 2 9B IT has 9.24B parameters in 42 layers. It uses grouped-query attention: 16 query heads share 8 key-value heads, so its KV cache is 2× smaller than full multi-head attention would need. It takes up to 8,192 tokens of context; one user at that length needs 2.63 GiB of KV cache at FP16 on top of the weights. Its repository is gated on Hugging Face, so LLMScale ships its architecture in a bundled database.

Gemma 2 9B IT memory by precision and context

One user, FP16 KV cache and vLLM's overhead on one GPU at its defaults for a card under 70 GB. Quantizing shrinks the weights, not the KV cache, so the saving narrows as the context grows.

Which single GPUs hold it

• BF16 (22.48 GiB): RTX 5090, Apple M4 Pro, A100 40GB, NVIDIA L40S
• FP8 (14.73 GiB): NVIDIA L4, Apple M4, NVIDIA A30, RTX 3090 / Ti
• INT4 (11.09 GiB): NVIDIA T4, RTX 4060 Ti 16GB, RTX 4070 Ti Super, RTX 4080 / Super

At 8k context for one user, the smallest cards first. A discrete card is budgeted at vLLM's default 92% of its memory, a unified-memory device at the share its GPU can address, and each carries vLLM's overhead at its own defaults.

How many users fit

Users at 8k context each, on one card. ✗: the weights leave no room for even one.

Fine-tuning Gemma 2 9B IT

Sequence length 2,048, batch size 1, gradient checkpointing on, Adam. Full fine-tuning keeps 16 bytes per parameter before activations; QLoRA stores the frozen base in 4-bit NF4.

Where the numbers come from

Weights are parameters × bytes per parameter. The KV cache is 2 × layers × key-value heads × head dimension × bytes × context × users. For Gemma 2 9B IT that is 336.0 KiB per token at FP16. 21 of its 42 layers attend over a sliding window of the last 4,096 tokens and hold no more than that, so past it each further token adds 168.0 KiB, the other 21 layers' share. These are capacity-planning figures: the serving engine — vLLM, TGI, llama.cpp — and its paged-attention efficiency still move the real number.

More on KV cache size and on GQA, MQA and MLA.

Work with Gemma 2 9B IT

• Size it for your workload
• Fine-tuning memory
• vLLM parameters
• Model anatomy
• Its config.json, explained
• What its files say about serving it

Other Gemma models

• Gemma 3 1B IT
• Gemma 3 4B IT
• Gemma 3 12B IT
• Gemma 2 27B IT
• Gemma 3 27B IT