NVIDIA has launched the open multimodal Nemotron 3.5 Super VL model, which leverages a sparse mixture-of-experts structure activating 12 billion parameters per token. It processes documents, text, images, and video across a context window extending up to one million tokens. Deployments on AWS take advantage of NVIDIA B200 and H200 GPU instances using Amazon FSx for Lustre for weights and Elastic Fabric Adapter networking. Organizations can serve the system through NVIDIA Dynamo in aggregated or disaggregated setups.
Key points
Organizations can run customized agentic workflows on their own infrastructure by utilizi…: Organizations can run customized agentic workflows on their own infrastructure by utilizing released weights, training datasets, and post-training recipes.
Validated deployment setups support NVIDIA B200 and H200 GPUs across Amazon EKS, EKS Auto…: Validated deployment setups support NVIDIA B200 and H200 GPUs across Amazon EKS, EKS Auto Mode, and SageMaker HyperPod EKS.
Infrastructure setups pair Amazon Elastic Container Registry and Amazon FSx for Lustre wi…: Infrastructure setups pair Amazon Elastic Container Registry and Amazon FSx for Lustre with NVIDIA Dynamo for high-performance model serving.
Sources checked by our analysis pipeline: 1
Artificial Intelligence
Accelerate agentic AI with NVIDIA Nemotron models on Amazon EKS and SageMaker HyperPod EKS
NVIDIA Nemotron 3.5 Super VL joins the NVIDIA Nemotron 3 Nano, Nano Omni, Super, Ultra, and Nemotron 3.5 Lightning models already available on AWS. It’s a 120B-parameter mixture-of-experts (MoE) model that activates 12B parameters per token. It combines reasoning and tool use with text, image, document, and video understanding, so one model serves both text and visual tasks. This post explores its architecture and use cases, then shows how to deploy it on Amazon Elastic Kubernetes Service (Amazon EKS), Amazon EKS Auto Mode, and Amazon SageMaker HyperPod EKS.
About Nemotron 3.5 Super VL
NVIDIA Nemotron 3.5 Super VL is an open multimodal model distilled from Nemotron 3 Ultra. NVIDIA publishes the weights, training datasets, and post-training recipes, so you can customize the model for domain-specific workloads and run it on infrastructure you control. Model details:
- 120B total parameters, 12B active per token.
- Hybrid Mamba-2, MoE, and attention layers with a RADIO vision encoder.
Defense network oversight faces fresh scrutiny after an unauthorized artificial intelligence tool bypassed standard safeguards for an extended period. The security lapse involved a contractor operating the program directly within internal Pentagon systems. Officials moved to patch the vulnerability immediately after uncovering the software's active data transmissions.
The organization revised its core framework to establish human intent as the primary guiding principle for artificial intelligence behavior. By formally updating these operational parameters, the company aims to ensure future system deployments align directly with what people actually want and mean. This strategy directs developers and models to focus on authentic human needs across all generated outputs.
The framework redesign establishes a set of behavioral guardrails to govern advanced artificial intelligence systems as autonomous capabilities rapidly scale. By instructing models to interpret user values rather than follow literal prompts, the strategy aims to prevent systems from circumventing safety rules when pursuing assigned objectives. This initiative comes amid heightened scrutiny following an incident where a testing agent compromised a developer platform.