NVIDIA Nemotron-3-Ultra-550B-A55B
serving recipe.

Floors are the minimum the vLLM recipe publishes across the whole replica, from vLLM's own runs. The control plane compares them with its own sizing from the model config.

• Tensor ParallelFits inside one node's interconnect. single_node_tp
• Multi-Node Tensor ParallelSpans nodes. multi_node_tp

• B200 · 8 GPUs

Wingsmith runs models through NVIDIA Dynamo, so it serves on NVIDIA GPUs. Other vendors are listed as the recipe states them.

vllm serve nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 \
  --served-model-name nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B \
  --trust-remote-code \
  --kv-cache-dtype fp8 \
  --max-num-seqs 16 \
  --max-model-len 262144 \
  --gpu-memory-utilization 0.90 \
  --max-num-batched-tokens 32768 \
  --enable-flashinfer-autotune \
  --async-scheduling \
  --mamba-backend triton \
  --mamba-ssm-cache-dtype float32 \
  --enable-auto-tool-choice \
  --tool-call-parser qwen3_coder \
  --reasoning-parser nemotron_v3 \
  --speculative-config '{"method":"mtp","num_speculative_tokens":5}'

Features

• Tool callingOn by defaultQwen3 Coder tool-call parser with automatic tool choice
• ReasoningOn by defaultNemotron v3 reasoning parser
• Speculative decodingOn by defaultMulti-Token Prediction with 5 speculative tokens

1. Register the weights

   In the console, or through the API. Resolving a slug proposes an artifact and writes nothing.

   curl "http://localhost:8000/v1/artifacts/resolve?source=hf:nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16"

2. Read the draft

   The draft carries this recipe as upstream evidence: the tested flags, the precision and memory floor, the parallelism layouts, and the hardware vLLM verified, with links back to the recipe. Nothing from it is applied without you, and the draft warns when the plane's own sizing and the published floor differ by a quarter or more.

3. Create the workload

   Choose the engine, device shape and scale. The plane picks the clusters, and each cluster's operator renders the deployment as a Dynamo graph. The one recipe default it applies, --max-num-seqs, is shown above.

4. Give it an endpoint

   One model name for clients, routed by the gateway to deployments on every cluster serving it.

Run the control plane locally →

• How much GPU memory does NVIDIA Nemotron-3-Ultra-550B-A55B need?The vLLM recipe publishes a minimum of 330 GB across the whole replica for its NVFP4 variant; BF16 1320 GB. That is vLLM's own measurement, not a per-GPU figure.
• Which parallelism layouts does the NVIDIA Nemotron-3-Ultra-550B-A55B recipe support?Tensor Parallel and Multi-Node Tensor Parallel.
• Can Wingsmith serve NVIDIA Nemotron-3-Ultra-550B-A55B?Wingsmith serves models through NVIDIA Dynamo on clusters with NVIDIA GPUs. Register nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 and the plane attaches this vLLM recipe to the draft as evidence: its tested flags, precision and memory floor, for you to review rather than applied on your behalf.
• Does NVIDIA publish a Dynamo recipe for NVIDIA Nemotron-3-Ultra-550B-A55B?Yes: 18 deployment entries for vLLM, aggregated and disaggregated. When a workload is placed, the plane defaults --max-num-seqs to the value the aggregated recipe publishes for that card, and names it on the draft.

From the vLLM recipe for nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 ↗ (snapshot at 4bcc189) and NVIDIA Dynamo's recipes at 165d854, Apache 2.0.