Showing posts with label GCP. Show all posts
Showing posts with label GCP. Show all posts

Tuesday, February 13, 2018

Google Cloud offers TensorFlow processors, delivering teraflops for ML

Cloud Tensor Processing Units (TPUs) are now available in beta on the Google Cloud Platform (GCP).

The Cloud TPUs are a family of Google-designed hardware accelerators that are optimized to speed up and scale up specific machine learning (ML) workloads programmed with TensorFlow.

Google's TPUs are built with four custom ASICs. Each Cloud TPU packs up to 180 teraflops of floating-point performance and 64 GB of high-bandwidth memory onto a single board. The boards can be connected via an ultra-fast network to form multi-petaflop ML supercomputers. These "TPU pods" will be available on GCP later this year.

The Cloud TPUs can be programmed with high-level, open source TensorFlow APIs. GCP is making a number of reference Cloud TPU model implementations available, including:

  • ResNet-50 and other popular models for image classification
  • Transformer for machine translation and language modeling
  • RetinaNet for object detection

Google adds that its cloud TPUs can also simplify the planning and management of ML computing resources. By purchasing the service, users benefit from the large-scale, tightly-integrated ML infrastructure that has been heavily optimized at Google over many years. The cloud TPUs are protected by the security mechanisms and practices that safeguard all Google Cloud services.

Google Builds a Custom ASIC for Machine Learning

Google has developed a custom ASIC for machine learning and artificial intelligence.

The Tensor Processing Unit (TPU) is tailored for TensorFlow, which is an open source software library for machine learning that was developed at Google.

In a blog posting,  Norm Jouppi, Distinguished Hardware Engineer at Google, discloses that the TPUs have already been in deployment in Google data centers for over a year, where they "deliver an order of magnitude better-optimized performance per watt for machine learning." The stealthy project to develop in-house silicon has been underway for several years.

A number of Google applications are already running on the Tensor Processing Units, including RankBrain, Street View, and the AlphaGo application that recently defeated the Go world champion, Lee Sedol.

Google plans to deliver machine learning as a service on its Google Cloud Platform by providing APIs for computer vision, speech, human language translation, etc.

https://cloudplatform.googleblog.com/2016/05/Google-supercharges-machine-learning-tasks-with-custom-chip.html


  • TensorFlow was originally developed by the Google Brain team and released under the Apache 2.0 open source license in November 2015. At the time, Google said TensorFlow can run on multiple CPUs and GPUs.

Tuesday, November 21, 2017

GCP trims pricing for cloud GPUs, SSDs

The Google Cloud Platform (GCP) is trimming the cost of NVIDIA Tesla GPUs attached to on-demand Google Compute Engine virtual machines by up to 36 percent. In US regions, each K80 GPU attached to a VM is priced at $0.45 per hour while each P100 costs $1.46 per hour.

This lowers the cost of running highly parallelized compute tasks on VMs and GPUs.

GCP is also lowering the price of preemptible Local SSDs by almost 40 percent compared to on-demand Local SSDs. In the US this means $0.048 per GB-month.


Sunday, December 18, 2016

Google Joins Cloud Foundry

The Google Cloud Platform (GCP) is now part of Cloud Foundry.

Google noted that it has been very active this year with the Cloud Foundry community, including the delivery of the BOSH Google CPI release, enabling the deployment of Cloud Foundry on GCP, and the recent release of the Open Service Broker API. The efforts have led to custom service brokers for eight of GCP services:

  • Google BigQuery
  • Google Cloud Storage
  • Google Cloud SQL
  • Google Cloud Pub/Sub
  • Google Cloud Vision API
  • Google Cloud Speech API
  • Google Cloud Natural Language API
  • Google Translation API


https://cloudplatform.googleblog.com/

See also