Phi 3 M3 Coder — Hardware Requirements & GPU Compatibility
ChatCode
15 downloads
48 likesSpecifications
- Publisher
- moelanoby
- Family
- Phi 3
- Parameters
- 3.8B
- Architecture
- Phi3ForCausalLM
- Context Length
- 4,096 tokens
- Vocabulary Size
- 32,064
- Release Date
- 2025-07-06
Get Started
HuggingFace
How Much VRAM Does Phi 3 M3 Coder Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| BF16 | 16.00 | 8.8 GB | 9.6 GB | 7.65 GB | Brain floating point 16 — preferred for training |
Which GPUs Can Run Phi 3 M3 Coder?
BF16 · 8.8 GBPhi 3 M3 Coder (BF16) requires 8.8 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 12+ GB is recommended. Using the full 4K context window can add up to 0.8 GB, bringing total usage to 9.6 GB. 28 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti, NVIDIA GeForce RTX 3080 Ti.
Runs great
— Plenty of headroom
Decent
— Enough VRAM, may be tightWhich Devices Can Run Phi 3 M3 Coder?
BF16 · 8.8 GB27 devices with unified memory can run Phi 3 M3 Coder, including NVIDIA DGX H100, NVIDIA DGX A100 640GB.
Runs great
— Plenty of headroomRelated Models
Frequently Asked Questions
- How much VRAM does Phi 3 M3 Coder need?
Phi 3 M3 Coder requires 8.8 GB of VRAM at BF16. Full 4K context adds up to 0.8 GB (9.6 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 3.8B × 16 bits ÷ 8 = 7.6 GB
KV Cache + Overhead ≈ 1.2 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 2 GB (at full 4K context)
VRAM usage by quantization
BF168.8 GBBF16 + full context9.6 GB- Can I run Phi 3 M3 Coder on a Mac?
Phi 3 M3 Coder requires at least 8.8 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 Phi 3 M3 Coder locally?
Yes — Phi 3 M3 Coder can run locally on consumer hardware. At BF16 quantization it needs 8.8 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Phi 3 M3 Coder?
At BF16, Phi 3 M3 Coder can reach ~333 tok/s on AMD Instinct MI300X. On NVIDIA GeForce RTX 4090: ~75 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 ÷ 8.8 × 0.55 = ~333 tok/s
Estimated speed at BF16 (8.8 GB)
AMD Instinct MI300X~333 tok/sNVIDIA GeForce RTX 4090~75 tok/sNVIDIA H100 SXM~249 tok/sAMD Instinct MI250X~206 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Phi 3 M3 Coder?
At BF16, the download is about 7.65 GB.
- Which GPUs can run Phi 3 M3 Coder?
28 consumer GPUs can run Phi 3 M3 Coder at BF16 (8.8 GB). Top options include AMD Radeon RX 6800, AMD Radeon RX 6800 XT, AMD Radeon RX 6900 XT, AMD Radeon RX 6700 XT. 17 GPUs have plenty of headroom for comfortable inference.
- Which devices can run Phi 3 M3 Coder?
27 devices with unified memory can run Phi 3 M3 Coder at BF16 (8.8 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.