README.md

---
license: other
license_name: nvidia-open-model-license
license_link: https://developer.download.nvidia.com/licenses/nvidia-open-model-license-agreement-june-2024.pdf
base_model: nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16
tags:
  - gguf
  - turboquant
  - kv-cache-quantization
  - nemotron
  - nvidia
  - mamba2
  - hybrid
  - moe
  - llama-cpp
  - quantized
library_name: gguf
pipeline_tag: text-generation
---

# Nemotron-3-Nano-30B-A3B-TurboQuant-GGUF-Q4_K_M

GGUF Q4_K_M weight-quantized variant of [nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16](https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16) optimised for use with **TurboQuant** KV cache compression via a dedicated llama.cpp fork.

> **Important:** TurboQuant KV cache types (`planar3`, `iso3`) are **not** available in upstream llama.cpp, standard Ollama, or LM Studio.
> They require a [specific llama.cpp fork](https://github.com/johndpope/llama-cpp-turboquant/tree/feature/planarquant-kv-cache).
> The GGUF file itself is a standard GGUF and works with any llama.cpp-compatible runtime using normal KV cache types (f16, q8_0, q4_0, etc.).

## Hardware compatibility

| Device | VRAM / RAM | Recommendation |
| --- | --- | --- |
| CPU host with ≥18 GB RAM | ~17.8 GB | works via llama.cpp; slower than GPU but no accelerator required |
| Apple Silicon (Metal) | ~19.4 GB | llama.cpp Metal backend; fast on M-series unified memory |
| NVIDIA GPU (partial offload) | split between GPU + RAM | offload as many layers as VRAM allows; rest on CPU |

## Overview

This model combines two independent compression techniques:

| Technique | What it does | Requirement |
|-----------|-------------|-------------|
| **GGUF Q4_K_M weight quantization** | Reduces model size from ~60 GB (BF16) to ~16.2 GB | Any llama.cpp-compatible runtime |
| **TurboQuant KV cache compression** — random rotation + Lloyd-Max scalar quantization (`--cache-type-k planar3 --cache-type-v planar3`) | Block-diagonal rotations / random rotation for compressed KV cache | [llama-cpp-turboquant fork](https://github.com/johndpope/llama-cpp-turboquant/tree/feature/planarquant-kv-cache) only |

## Quickstart

### Option A — With TurboQuant KV cache (fork required)

You must build from the TurboQuant-enabled llama.cpp fork:

```bash
# Clone and build the fork
git clone https://github.com/johndpope/llama-cpp-turboquant.git
cd llama-cpp-turboquant && git checkout feature/planarquant-kv-cache

# CUDA (Windows/Linux)
cmake -B build -DGGML_CUDA=ON -DCMAKE_BUILD_TYPE=Release && cmake --build build -j

# Metal (Apple Silicon)
cmake -B build -DGGML_METAL=ON -DGGML_METAL_EMBED_LIBRARY=ON -DCMAKE_BUILD_TYPE=Release && cmake --build build -j

# Run with TurboQuant KV cache
./build/bin/llama-cli -m Nemotron-3-Nano-30B-A3B-TurboQuant-GGUF-Q4_K_M.gguf \
  --cache-type-k planar3 --cache-type-v planar3 \
  -ngl 99 -fa \
  -p "Explain quantum computing"

# Or run as a server
./build/bin/llama-server -m Nemotron-3-Nano-30B-A3B-TurboQuant-GGUF-Q4_K_M.gguf \
  --cache-type-k planar3 --cache-type-v planar3 \
  -ngl 99 -fa --jinja
```

### Option B — With standard llama.cpp / LM Studio / Ollama

The GGUF works as a normal quantised model. You won't get TurboQuant-specific KV cache benefits, but standard KV cache quantization (q8_0, q4_0) still reduces VRAM significantly.

**llama.cpp (upstream)**
```bash
llama-cli -m Nemotron-3-Nano-30B-A3B-TurboQuant-GGUF-Q4_K_M.gguf \
  --cache-type-k q8_0 --cache-type-v q8_0 \
  -ngl 99 -fa \
  -p "Explain quantum computing"
```

**LM Studio**
1. Download the GGUF file and load in LM Studio.
2. Enable **Developer Mode** (Settings → Developer).
3. In the model loader's advanced settings, set **Flash Attention** to ON.
4. Set **K Cache Quantization** and **V Cache Quantization** to `q8_0` (or `q4_0` for more aggressive VRAM savings).
5. Note: LM Studio does not currently support TurboQuant's `planar3` cache types. Track [this feature request](https://github.com/lmstudio-ai/lmstudio-bug-tracker/issues/1719) for updates.

**Ollama**
```bash
# Standard Ollama does not support TurboQuant cache types.
# Use with default or q8_0 KV cache via OLLAMA_KV_CACHE_TYPE=q8_0
OLLAMA_KV_CACHE_TYPE=q8_0 OLLAMA_FLASH_ATTENTION=1 ollama run majentik/Nemotron-3-Nano-30B-A3B-TurboQuant-GGUF-Q4_K_M
```

## Specifications

| Property | Value |
|----------|-------|
| Base Model | [nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16](https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16) |
| Architecture | Mamba-2 + Transformer hybrid Sparse MoE |
| Parameters | 30.7B total, 3.2B active per token |
| Context Length | 1M |
| Weight Quantization | GGUF Q4_K_M (popular 4-bit, best quality/size tradeoff) |
| Original Size (BF16) | ~60 GB |
| Quantized File Size | ~16.2 GB |
| KV Cache (TurboQuant) | 3-bit via `--cache-type-k planar3 --cache-type-v planar3` (fork only) |
| KV Cache (standard) | q8_0, q4_0, f16, etc. (any llama.cpp runtime) |
| License | other |
| Modalities | Text only |
| Compatible Runtimes | llama.cpp, LM Studio, Ollama, koboldcpp |

## What is TurboQuant?

[TurboQuant](https://arxiv.org/abs/2504.19874) (ICLR 2026) is a KV cache compression method that applies a random orthogonal rotation followed by optimal scalar quantization. Bit-identical prefill logits at 4-bit on tested models, with up to 4-8× memory savings for long sequences.

**Benchmarks from the TurboQuant repository** (Llama 3.1 8B, RTX 5090 — results will vary by model and hardware):

| Metric | TurboQuant (4-bit) | Standard q4_0 |
|--------|--------------------|---------------|
| Quality | Bit-identical prefill | Lossy |
| KV Compression | ~4× vs FP16 | ~4× vs FP16 |
| Speedup (Apple Silicon) | 1.4–1.7× | — |

> **Note:** These benchmarks are from the TurboQuant repository using Llama 3.1 8B on an RTX 5090. Performance on Nemotron-3-Nano-30B-A3B will differ. Independent benchmarks for this specific model are welcome — please open a discussion if you have results to share.

## Current Status of TurboQuant in the Ecosystem

| Runtime | TurboQuant Support | Standard KV Quant |
|---------|---------------------|-------------------|
| llama.cpp (upstream) | ❌ Not merged | ✅ q8_0, q4_0, q4_1, iq4_nl, q5_0, q5_1 |
| llama-cpp-turboquant fork | ✅ planar3 | ✅ All standard types |
| LM Studio | ❌ [Requested](https://github.com/lmstudio-ai/lmstudio-bug-tracker/issues/1719) | ✅ Via advanced settings |
| Ollama | ❌ Not supported | ✅ Via OLLAMA_KV_CACHE_TYPE |
| koboldcpp | ❌ Not supported | ✅ Standard types |

## Recommended Settings

For VRAM-constrained setups, standard q8_0 KV cache quantization already halves KV cache memory with negligible quality impact. Flash Attention should always be enabled — it is required for V cache quantization and improves memory efficiency regardless.

| VRAM | Suggested Configuration |
|------|------------------------|
| 24 GB (RTX 4090) | Q4_K_M + q8_0 KV cache + Flash Attention, 8K–16K context |
| 16 GB | Q4_K_M + q4_0 KV cache + Flash Attention, 4K–8K context |
| 48+ GB | Q4_K_M + f16 KV cache, full 32K+ context |

## See Also

- [RotorQuant GitHub](https://github.com/scrya-com/rotorquant)
- [llama-cpp-turboquant fork](https://github.com/johndpope/llama-cpp-turboquant/tree/feature/planarquant-kv-cache)
- [TurboQuant llama.cpp discussion](https://github.com/ggml-org/llama.cpp/discussions/20969)
- [TurboQuant paper (arXiv: 2504.19874)](https://arxiv.org/abs/2504.19874)
- [Base model: nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16](https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16)
- [Nemotron-3-Nano-30B-A3B announcement](https://huggingface.co/blog/nvidia/nemotron-3-nano-efficient-open-intelligent-models)