Google Cloud announced two Dataflow changes aimed at large-scale AI and analytics workloads: general availability of pause and resume for batch jobs, and support for G4 virtual machines powered by NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs.
The pause and resume capability is intended to improve compute efficiency for long-running batch processing. Customers can suspend work when resources are not needed and continue later, potentially reducing idle infrastructure use for workloads that do not require uninterrupted execution.
GPU support expands the hardware options available to Dataflow users running inference and other accelerated processing tasks. Google presented the changes as part of its effort to make data preparation, model evaluation and inference easier to operate without customers managing the underlying infrastructure.
The announcement places Dataflow more directly within Google Cloud's AI infrastructure portfolio. It combines a serverless data-processing service with access to newer accelerator hardware, while retaining managed autoscaling and pipeline operations.
The release is a product update rather than a change to Google Cloud's overall pricing or regional footprint, but it is relevant to customers balancing accelerator performance against utilization and operational cost.
