By Dave Salvator, Senior Manager, Product Marketing at NVIDIA NVIDIA and Google Cloud are making it possible for applications to push the boundaries of accelerated AI across a wide array of applications. With its new A2 VM, announced today, Google Cloud provides customers the largest configuration of 16 NVIDIA A100 GPUs in a single VM. Also […]
By Dave Salvator, Senior Manager, Product Marketing at NVIDIA
NVIDIA and Google Cloud are making it possible for applications to push the boundaries of accelerated AI across a wide array of applications. With its new A2 VM, announced today, Google Cloud provides customers the largest configuration of 16 NVIDIA A100 GPUs in a single VM. Also available are smaller GPU configurations including 1, 2, 4, and 8 GPUs per VM for added flexibility. The single VM offering features NVIDIA’s NVLink Fabric to deliver greater multi-GPU scalability for ultrafast deep learning training.
Additionally, the A2 is supported by many Google Managed Services for developers:
- Google Kubernetes Engine (GKE)
- Google Cloud AI Platform training
- Dataproc – fully managed service for SPARK and DASK
- Support for A100 with pre-built VM images including versions for CUDA, TensorFlow, PyTorch, and RAPIDS
Developers can take advantage of NVIDIA NGC, a repository for containerized, pre-configured software to accelerate A100 development workflows and building solutions in Google Cloud.
Google Cloud was the first to announce their A100 private alpha program and now has referenceable customers including Cash App, which uses NVIDIA A100 GPUs to power mobile payment innovation and research. Google Cloud AI and NVIDIA delivered a 66% speedup to the processing time needed for Cash App to complete a critical machine learning workflow.
Head over to Google Cloud’s GPU Instances page to learn more about the new A2 instance.