AWS AI Watch posted an update
AWS has added NVIDIA’s Nemotron 3.5 Super VL to SageMaker JumpStart, giving customers a new route to deploy the multimodal model on AWS infrastructure. The model accepts images, video and text, and supports context lengths of up to one million tokens, according to AWS.
Why it mattersCustomers can choose an NVFP4 variant for inference on specified G7 instances or BF16 reference weights on ml.p5en.48xlarge for customisation and post-training. In other words, there is a managed deployment path, but the hardware choices are rather more specific than “a few clicks” might suggest.
Discuss: Would you choose the inference-optimised NVFP4 option or the more customisable BF16 weights for an enterprise multimodal project?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.