Software-defined networking meets HPC

Опубликовано 15 сентября 2020, 16:18

Google Cloud has the best networking infrastructure that any cloud can offer. Google Cloud was built on the same infrastructure that delivers services like Google Search and other hyperscale services. HPC applications and workloads can leverage the same high-performing network to deliver timely insights into the ever increasing deluge of data and constant need of computing power. Explore how Google’s software-defined network infrastructure can meet the demanding requirements like low latency, low jitter, scale, high throughput, and elasticity for complex scientific computational applications.

In many HPC workloads, achieving the best end-to-end performance of an application requires configuring your network stack to perform optimally for MPI or other communication protocols. Learn how customers can run MPI-enabled applications and provide a roadmap for high throughput virtual machines (general purpose, GPU, TPU, Google Kubernetes Engine).


Speakers: Manoj Jayadevan, Parveen Patel

Watch more:
Google Cloud Next ’20: OnAir → goo.gle/next2020

Subscribe to the GCP Channel → goo.gle/GCP

#GoogleCloudNext

NET203

event: Google Cloud Next 2020; re_ty: Publish; product: Cloud - Containers - Google Kubernetes Engine (GKE); fullname: Manoj Jayadevan, Parveen Patel;