NVIDIA Nemotron 3 Super 120B A12B (Q2_K_XL) — 74.2 GBon NVIDIA H100
Overview
NVIDIA Nemotron 3 Super 120B A12B is a 123.61B parameter moe language model by NVIDIA, with code, multilingual, thinking, tool-calls capabilities. It supports a context window of up to 262,144 tokens.
Nemotron 3 Super 120B A12B is a 123.61-billion-parameter hybrid Mamba-2 Transformer LatentMoE model from NVIDIA, activating 12 billion parameters per token across 22 of 512 routed experts plus 1 shared expert. Trained on over 25 trillion tokens, it targets agentic reasoning, code generation, tool calling, and multilingual conversation in 7 languages. A 256K context window, toggleable thinking mode, and multi-token prediction enable high-throughput inference for complex multi-agent workflows. Its MoE sparsity quantizes well to GGUF for self-hosted deployment on multi-GPU setups.
At Q2_K_XL quantization (low quality tier), the model weighs 50.9 GB. This fits within the 80 GB of VRAM on NVIDIA H100, enabling full GPU inference.
The NVIDIA H100 80 GB is a datacenter GPU with 80 GB of HBM3 VRAM and 3350 GB/s memory bandwidth. It delivers 1979 FP16 TFLOPS on the Hopper architecture, making it the fastest single-GPU option for large language model inference. It handles models up to 70B parameters with high throughput. Built for datacenter teams running demanding production AI workloads.
Hardware Requirements
| Model size | 50.9 GB |
| VRAM available | 80 GB |
| VRAM used | 74.2 GB |
| System RAM | |
| Min RAM required | 0 GB |
| GPU layers | 88 / 88 |
| Context size | 262,144 |
| Backend | cuda13 |
| Flash attention | Yes |
| Reading from disk | Yes |
Performance Notes
Deploy
Prerequisites
Ensure your GPU nodes are prepared with the NVIDIA container toolkit:
ansible-playbook prositronic.infra.nvidia_container_toolkitCommand
helmfile --state-values-file <(curl -s https://www.prositronic.eu/values/nemotron-3-super-120b-a12b/q2_k_xl/nvidia-h100-80gb.yaml) applyGenerated values.yaml
/values/nemotron-3-super-120b-a12b/q2_k_xl/nvidia-h100-80gb.yaml
Loading values…Frequently Asked Questions
How much VRAM does NVIDIA Nemotron 3 Super 120B A12B (Q2_K_XL) need?
The Q2_K_XL quantization of NVIDIA Nemotron 3 Super 120B A12B requires 50.9 GB. All 88 layers fit in the 80 GB of VRAM available on NVIDIA H100, enabling full GPU acceleration.
Can I run NVIDIA Nemotron 3 Super 120B A12B on NVIDIA H100?
Yes. NVIDIA H100 provides 80 GB of VRAM, which is enough to run NVIDIA Nemotron 3 Super 120B A12B (Q2_K_XL) with all layers on the GPU for optimal performance.
What is quantization?
Quantization reduces a model's numerical precision from its original floating-point format to a more compact representation. This shrinks the file size and VRAM footprint, making it possible to run large models on consumer hardware. The trade-off is a small reduction in output quality. Q2_K_XL compresses NVIDIA Nemotron 3 Super 120B A12B from its original size down to 50.9 GB.
What quantization should I choose for NVIDIA Nemotron 3 Super 120B A12B?
Q2_K_XL is a low-quality quantization. Higher-quality quants (Q8, Q6) preserve more model accuracy but need more VRAM. Lower quants (Q4, Q3, Q2) reduce VRAM usage at the cost of some quality. Choose based on your available hardware and quality requirements.
What is flash attention and why is it enabled?
Flash attention is a memory-efficient algorithm that speeds up the attention mechanism in transformer models. It reduces VRAM usage during inference by avoiding the materialisation of the full attention matrix. For NVIDIA Nemotron 3 Super 120B A12B on NVIDIA H100, flash attention is enabled to maximise context length and throughput within the available 80 GB of VRAM.
What is MoE and how does it affect deployment?
NVIDIA Nemotron 3 Super 120B A12B uses a Mixture-of-Experts (MoE) architecture with 512 experts, of which 22 are active per token. This means only a fraction of the model weights are used for each inference step, allowing MoE models to be larger in total parameter count while remaining efficient at inference time.