V LLMVRAMCalc
EN中文ESDE
Calculator › Nemotron 3 Ultra 550B-A55B › B200 180GB

Nemotron 3 Ultra 550B-A55B on B200 180GB (180GB) — VRAM, KV Cache & Throughput

Nemotron 3 Ultra 550B-A55B (550B total, 55B active, Hybrid Linear/Mamba + MoE) uses 310.97 GiB of 4-bit weights (Q4_K_M), requiring 311.95 GiB total at 8K context and 312.23 GiB at 32K on B200 180GB (~3.8 tok/s decode).

Interactive Sizing Workbench

Full Compatibility Report →
180 GB · 8000 GB/s
Exceeds Available GPU Memory
Config-derived formula HF config.json ↗
Memory Allocation Breakdown (GiB) 311.95 GiB (334.95 GB) / 180 GiB (173%)
Model Weights: 310.97 GiB KV Cache & State: 0.41 GiB Runtime & Activation Buffers: 0.57 GiB System RAM Offload: 0 GiB
Required GPU Memory
311.95 GiB
334.95 GB (decimal)
Usable Hardware Memory
171 GiB
180 GB (physical)
Safe Headroom
-140.95 GiB
173% utilization
Est. Decode Throughput
~3.8 tok/s
55B active params
Est. Prefill Throughput
~10636 tok/s
4K Prompt TTFT: ~385 ms
Est. Cost / 1M Tokens
$874.27
Hybrid Linear/Mamba + MoE · NVIDIA Open

Recommended Hardware for Current Configuration

Best Single GPU
Multi-GPU Required
—
Best Value / Used GPU
2x B200 180GB
360GB · ~$80,000
Apple Unified Memory
Mac Studio M3 Ultra 512GB
~$9,500
Cloud From
RunPod · $11.96 / hr
2x B200 180GB

Production Launch Commands

llama.cpp (GGUF) 
llama-server -hf nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16:Q4_K_M \
  --ctx-size 8192 -ngl 999 --flash-attn
vLLM 
vllm serve nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 \
  --max-model-len 8192 --gpu-memory-utilization 0.92
Live Token Speed Simulator (TTFT ~385ms · ~3.8 tok/s)
[Nemotron 3 Ultra 550B-A55B · 311.95 GiB] allocated on target GPU. Prefill completed in TTFT window; streaming autoregressive tokens at memory-bandwidth-bound rate...

Cloud GPU Comparison & Rent-vs-Buy TCO

Verified public rates across RunPod, Vast.ai, Vultr, and DigitalOcean.

Verified: 2026-11-04
Purchase break-even
~27.9 months
Est. local power: $21.28 / mo
Provider Recommended Instance Total VRAM Hourly Rate Monthly at Selected Hours Billing Mode Deploy
RunPod 2x B200 180GB (Community Cloud) 360 GB $11.96 / hr $1454.34 / mo Per-second Deploy GPU ↗
Vast.ai 2x B200 180GB (Verified Host) 360 GB $15.00 / hr $1824.49 / mo Spot / On-demand Deploy GPU ↗
Vultr ($300 Free Credit*) 8x A100 80GB SXM (Dedicated Instance · $300 Credit*) 640 GB $19.18 / hr $2331.80 / mo Hourly Deploy GPU ↗
Vultr Cloud GPU 8x A100 80GB SXM (Dedicated Instance) 640 GB $19.18 / hr $2331.80 / mo Hourly Deploy GPU ↗
DigitalOcean 8x Instinct MI300X 192GB (Dedicated Instance) 1536 GB $20.72 / hr $2519.55 / mo Per-second Deploy GPU ↗

Rates reflect public listings on the verification date and fluctuate with regional inventory.

Fine-Tuning VRAM Requirements (QLoRA / LoRA / Full) — Nemotron 3 Ultra 550B-A55B

QLoRA (4b) Exceeds Available GPU Memory
319.87 GiB
Model Weights: 310.97 GiB · Gradients + Optimizer: 7.7 GiB · Act: 1.2 GiB
Recommended GPU: 2x B200 180GB (360GB)
LoRA (16b) Exceeds Available GPU Memory
1038.17 GiB
Model Weights: 1024.45 GiB · Gradients + Optimizer: 12.1 GiB · Act: 1.62 GiB
Recommended GPU: 8x B200 180GB (1440GB)
Full BF16 Exceeds Available GPU Memory
4203.1 GiB
Model Weights: 1024.45 GiB · Gradients + Optimizer: 3175.8 GiB · Act: 2.85 GiB
Recommended GPU: Multi-GPU Required

Engineering Sizing Notes & Optimization Playbook

Quick Hardware Verdict

A single B200 180GB (180GB) cannot hold 4-bit Nemotron 3 Ultra 550B-A55B (311.95 GiB at 8K) entirely in VRAM. Use multi-GPU tensor parallelism, offload layers to system RAM, or deploy on a larger cloud GPU.

KV Cache & Runtime Tuning

Nemotron 3 Ultra 550B-A55B uses Hybrid Linear/Mamba + MoE (550B total, 55B active per token). Moving from 8K to 32K context expands FP16 KV cache from 0.41 GiB to 0.69 GiB; enabling 8-bit KV cache (llama.cpp -ctk q8_0 -ctv q8_0 or vLLM --kv-cache-dtype fp8) saves 0.19 GiB at 32K (reducing total VRAM to 312.04 GiB).

VRAM Scaling Across Context Windows (4K to 128K, 4-bit Weights)

Nemotron 3 Ultra 550B-A55B VRAM scaling from 4K to 128K context (4K: 311.89 GiB, 8K: 311.95 GiB, 32K: 312.23 GiB, 128K: 313.35 GiB) 4K 311.89 GiB (KV 0.36 GiB) 8K 311.95 GiB (KV 0.41 GiB) 32K 312.23 GiB (KV 0.69 GiB) 128K 313.35 GiB (KV 1.81 GiB) B200 180GB (180GB)
Model weights remain constant while attention KV cache grows with context length and batch size.

Quantization Precision & Memory Matrix — Nemotron 3 Ultra 550B-A55B @ 8K

Quantization Format Weights (GiB) KV Cache Total VRAM (GiB) Decode Speed Compatibility (B200 180GB) Recommended GPU Data Source
Native Weights (BF16 / FP8 / MXFP4 — Lossless) (BF16) 1024.45 GiB 0.41 GiB 1025.43 GiB (1101.05 GB) ~0.8 tok/s Exceeds Available GPU Memory Multi-GPU Required Official checkpoint size
FP8 / GGUF Q8_0 (8-bit — Near-Lossless) (Q8_0) 544.24 GiB 0.41 GiB 545.22 GiB (585.43 GB) ~1.4 tok/s Exceeds Available GPU Memory Multi-GPU Required Config-derived formula
GGUF Q6_K (6-bit — High Precision) (Q6_K) 420.03 GiB 0.41 GiB 421.01 GiB (452.06 GB) ~2.1 tok/s Exceeds Available GPU Memory Multi-GPU Required Config-derived formula
GGUF Q5_K_M (5-bit — Balanced) (Q5_K_M) 364.54 GiB 0.41 GiB 365.52 GiB (392.47 GB) ~2.7 tok/s Exceeds Available GPU Memory Multi-GPU Required Config-derived formula
Q4_K_M / UD-Q4_K_XL / AWQ (4-bit — Recommended) (Q4_K_M) 310.97 GiB 0.41 GiB 311.95 GiB (334.95 GB) ~3.8 tok/s Exceeds Available GPU Memory Multi-GPU Required Config-derived formula
GGUF IQ4_XS (4.25-bit — Compact) (IQ4_XS) 272.43 GiB 0.41 GiB 273.41 GiB (293.57 GB) ~5.2 tok/s Exceeds Available GPU Memory Multi-GPU Required Config-derived formula
GGUF Q3_K_M (3-bit — Low VRAM) (Q3_K_M) 250.62 GiB 0.41 GiB 251.6 GiB (270.15 GB) ~6.8 tok/s Exceeds Available GPU Memory Instinct MI355X 288GB (288GB) Config-derived formula
GGUF UD-Q2_K_XL (2-bit — Extreme MoE) (UD-Q2_K_XL) 176.52 GiB 0.41 GiB 177.5 GiB (190.59 GB) ~147.3 tok/s Tight Fit (<8% Headroom) Instinct MI325X 256GB (256GB) Config-derived formula

Technical Specifications & FAQ

Does Nemotron 3 Ultra 550B-A55B fit on B200 180GB (180GB)?

In 4-bit (Q4_K_M) at 8,192 context, Nemotron 3 Ultra 550B-A55B requires 311.95 GiB total GPU memory (310.97 GiB weights + 0.41 GiB KV cache + 0.57 GiB runtime buffers) against 171 GiB usable on B200 180GB, delivering ~3.8 tok/s decode.

How much VRAM does Nemotron 3 Ultra 550B-A55B need at 32K context?

At 32,768 context tokens, Nemotron 3 Ultra 550B-A55B (Hybrid Linear/Mamba + MoE) allocates 0.69 GiB for KV cache and state, bringing total 4-bit VRAM to 312.23 GiB (335.25 GB decimal).

Evaluate This Model on Other GPUs

RTX 3060 12GB (12GB)RTX 4060 Ti 16GB (16GB)RTX 4070 Ti Super (16GB)RTX 4080 Super (16GB)RTX 5070 12GB (12GB)RTX 5070 Ti (16GB)RTX 5080 (16GB)RTX 3090 (24GB)RTX 4090 (24GB)RTX 4090 D (24GB)RTX 5090 (32GB)RTX 5090 D (32GB)RTX A6000 (48GB)RTX 6000 Ada (48GB)RTX PRO 6000 Blackwell (96GB)DGX Spark (GB10 128GB) (128GB)L4 24GB (24GB)A40 48GB (48GB)L20 48GB (48GB)L40S 48GB (48GB)A800 80GB SXM (80GB)A100 80GB SXM (80GB)H800 80GB SXM (80GB)H100 80GB SXM (80GB)H20 96GB HBM3 (96GB)H20 141GB HBM3e (141GB)H200 141GB SXM (141GB)B300 288GB Blackwell Ultra (288GB)Ascend 910B 64GB (昇腾) (64GB)Ascend 910C 128GB (昇腾) (128GB)Atlas 300I Duo 96GB (310P) (96GB)MLU590 80GB (寒武纪思元) (80GB)DCU K100-AI 64GB (海光深算) (64GB)MetaX C500 64GB (沐曦曦云) (64GB)Kunlunxin P800 64GB (昆仑芯) (64GB)MTT S4000 48GB (摩尔线程) (48GB)Radeon RX 7900 XTX 24GB (24GB)Radeon AI PRO R9700 32GB (32GB)Ryzen AI Max+ 395 128GB (128GB)Instinct MI300X 192GB (192GB)Instinct MI325X 256GB (256GB)Instinct MI355X 288GB (288GB)Arc Pro B60 24GB (24GB)Gaudi 3 128GB HBM2e (128GB)Mac mini M4 Pro 64GB (64GB)MacBook Pro M4 Max 128GB (128GB)Mac Studio M4 Ultra 256GB (256GB)Mac Studio M3 Ultra 512GB (512GB)

Evaluate Other Models on This Hardware

Qwen3.8 Flash Next (125B-A6B)Qwen3.8 27BQwen3.8 2.4T-A95BQwen3.6 35B-A3BQwen3.5 397B-A17BQwen3.5 122B-A10BQwen3.5 35B-A3BQwen3.5 27BQwen3.5 9BQwen3.5 4BQwen3 Coder 480B-A35BQwen3 Coder Next 80B-A3BQwen3 Coder 30B-A3BQwen3 30B-A3B Instruct 2507Qwen3 235B-A22B Instruct 2507Qwen3 32BQwen3 14BQwen3 8BQwen3 4B Instruct 2507Qwen3 1.7BQwen3 0.6BQwen2.5 72B InstructQwen2.5 Coder 32B InstructQwQ 32B ReasoningDeepSeek V4.1 Flash (552B-A16B)DeepSeek V4 FlashDeepSeek V3.2 (685B)DeepSeek R1-0528 671BDeepSeek V3-0324 671BDeepSeek R1 Distill Llama 70BDeepSeek R1 Distill Qwen 32BDeepSeek R1 Distill Qwen 14BDeepSeek R1 Distill Llama 8Bgpt-oss-20bgpt-oss-120bGemma 4 31B ITGemma 4 26B-A4B ITGemma 4 12B ITGemma 3 27B ITGemma 3 12B ITGemma 3 4B ITGLM-5.3 Flash (320B-A18B)GLM-5.3 (753B)GLM-5 (744B)GLM-4.5 (355B-A32B)GLM-4.5 Air (110B-A12B)GLM-4 32B-0414Kimi K3 (2.8T-A104B)Kimi K2 Instruct (1T-A32B)Kimi-Dev 72BMiniMax M3.1 Flash (428B-A23B)MiniMax M2.5 (229B-A10B)Step-3.7 Flash (198B-A11B)Step-3.5 Flash (197B-A11B)ERNIE 4.5 300B-A47B (文心)ERNIE 4.5 21B-A3B (文心)Hunyuan A13B Instruct (混元 80B)MiniCPM5 2B (面壁)Llama 4 Maverick 17B-128E (400B)Llama 4 Scout 17B-16E (109B)Llama 3.1 405B InstructLlama 3.3 70B InstructLlama 3.1 8B InstructLlama 3.2 3B InstructMistral Large 4 (1T-A52B)Mistral Large 2 123BMixtral 8x22B InstructMixtral 8x7B InstructMistral Small 3.2 24BDevstral Small 24BNemotron 3 Super 120B-A12BNemotron 3 Nano 30B-A3BCohere Command A 111BPhi-4 14BLiquid LFM3 1.7B