Whether your an ML researcher, infrastructure manager or HPC application owner, you probably want to get started with GPU infrastructure quickly, run comfortably in production and dynamically scale as needed.

We will give a wide overview of are various GPU offerings and features that are often used with ML and HPC workloads. From there, we will discuss real-world customer story of how they manage their GPU compute infrastructure on GCP.

Some things we will discuss: Our various GPU products, including the new NVIDIA Tesla T4 and V100 GPU. Custom VM shapes with GPUs, Deep Learning VM Image for quickly getting started. Dynamic Scaling. Preemptible GPUs for low cost, batch workloads. GPU integration with GKE.

Build with Google Cloud → bit.ly/2TVEpZx

Watch more:
Next '19 Compute Sessions here → bit.ly/Next19Compute
Next ‘19 All Sessions playlist → bit.ly/Next19AllSessions

Subscribe to the GCP Channel → bit.ly/GCloudPlatform

Speaker(s): Chris Kleban

Session ID: CMP103
product: Cloud - AI and Machine Learning - AI Infrastructure; fullname: Chris Kleban; event: Google Cloud Next 2019;