Llama 3.2 3B Base — Hardware Requirements & GPU Compatibility
Chat
36 downloads
2 likesSpecifications
- Publisher
- 1024m
- Family
- Llama 3
- Parameters
- 3.2B
- Architecture
- LlamaForCausalLM
- Context Length
- 131,072 tokens
- Vocabulary Size
- 128,256
- Release Date
- 2024-09-25
- License
- llama3.2
Get Started
HuggingFace
How Much VRAM Does Llama 3.2 3B Base Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| BF16 | 16.00 | 7.0 GB | 21.8 GB | 6.43 GB | Brain floating point 16 — preferred for training |
Which GPUs Can Run Llama 3.2 3B Base?
BF16 · 7.0 GBLlama 3.2 3B Base (BF16) requires 7.0 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 10+ GB is recommended. Using the full 131K context window can add up to 14.8 GB, bringing total usage to 21.8 GB. 35 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti, NVIDIA GeForce RTX 3080.
Runs great
— Plenty of headroom
Which Devices Can Run Llama 3.2 3B Base?
BF16 · 7.0 GB33 devices with unified memory can run Llama 3.2 3B Base, including NVIDIA DGX H100, NVIDIA DGX A100 640GB, MacBook Air 13" M3 (8 GB).
Runs great
— Plenty of headroomDecent
— Enough memory, may be tightRelated Models
Frequently Asked Questions
- How much VRAM does Llama 3.2 3B Base need?
Llama 3.2 3B Base requires 7.0 GB of VRAM at BF16. Full 131K context adds up to 14.8 GB (21.8 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 3.2B × 16 bits ÷ 8 = 6.4 GB
KV Cache + Overhead ≈ 0.6 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 15.4 GB (at full 131K context)
VRAM usage by quantization
BF167.0 GBBF16 + full context21.8 GB- Can I run Llama 3.2 3B Base on a Mac?
Llama 3.2 3B Base requires at least 7.0 GB at BF16, which exceeds the unified memory of most consumer Macs. You would need a Mac Studio or Mac Pro with a high-memory configuration.
- Can I run Llama 3.2 3B Base locally?
Yes — Llama 3.2 3B Base can run locally on consumer hardware. At BF16 quantization it needs 7.0 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Llama 3.2 3B Base?
At BF16, Llama 3.2 3B Base can reach ~419 tok/s on AMD Instinct MI300X. On NVIDIA GeForce RTX 4090: ~94 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: AMD Instinct MI300X → 5300 ÷ 7.0 × 0.55 = ~419 tok/s
Estimated speed at BF16 (7.0 GB)
AMD Instinct MI300X~419 tok/sNVIDIA GeForce RTX 4090~94 tok/sNVIDIA H100 SXM~313 tok/sAMD Instinct MI250X~259 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Llama 3.2 3B Base?
At BF16, the download is about 6.43 GB.
- Which GPUs can run Llama 3.2 3B Base?
35 consumer GPUs can run Llama 3.2 3B Base at BF16 (7.0 GB). Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT, AMD Radeon RX 7600. 27 GPUs have plenty of headroom for comfortable inference.
- Which devices can run Llama 3.2 3B Base?
33 devices with unified memory can run Llama 3.2 3B Base at BF16 (7.0 GB), including Mac Mini M4 (16 GB), Mac Mini M4 (32 GB), Mac Mini M4 Pro (24 GB), Mac Mini M4 Pro (48 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.