Не пропустите эти только что объявленные сессии для Cloud OnAir: Путешествие с больших данных в AI. Независимо от того, являетесь ли вы новичком в искусственном интеллекте или опытным профессионалом, эти углубленные сессии могут помочь вам преобразовать данные в реальные идеи.
Вот краткий взгляд на некоторые из 20 + сеансов, которые будут показаны.
Big Data track — Узнайте, как быстро и уверенно погрузиться в свои данные с 30-минутными сеансами, такими как:

• Хранилище данных Reimagine с помощью Google BigQuery: что нового и что нового?
• Интеллектуальный Интернет вещей: начните анализ ваших глобальных данных устройства для анализа в реальном времени
• Zen Руководство по подготовке данных для аналитики: обучение и передовая практика

AI track — получите представление о том, как заставить AI работать для вас, в том числе:

• Курс краха API-интерфейсов Google Cloud Machine Learning
• Переосмысление беседы с облачной речью и диалоговым потоком

Ускоренное машинное обучение в облаке: графические процессоры, облачные TPU и многое другое
Чат в реальном времени с экспертами Google Cloud на этом глобальном онлайн-мероприятии, чтобы узнать, как ИИ может разблокировать мощь ваших данных. Следите за полным списком сессий и не забудьте зарегистрироваться сегодня.
https://cloudonair.withgoogle.com/events/​data-to-ai

cloudplatform.googleblog.com/2017/08/introducing-Network-Service-Tiers-your-cloud-network-your-way.html

New Cloud Regions for all.
GCP now serves over a billion end users. To support customer growth, we’re adding eight new regions in 2017—Mumbai, Singapore, Sydney, Northern Virginia, São Paulo, London, Finland, and Frankfurt. Regional expansion benefits all with lower latency and cost, improved control over data localization, and enhanced stability and performance.
cloud.google.com/about/locations/
True multi-cloud support. No lock-in.
Container Engine (GKE) now delivers true cloud federation. Powered by newly released Kubernetes 1.4, GKE automated container management lets you easily deploy across multiple clusters—and multiple clouds. New capabilities include monitoring cluster add-ons, one-click cluster spin-up, and support for the new Google Container-VM Image (GCI).
cloud.google.com/container-engine/
Move to cloud data warehousing.
With new capabilities core to common large-scale business workloads, BigQuery is now a full-fledged data warehouse. Unlimited flat-rate pricing—and unlimited queries—make data storage costs predictable. And with support for standard SQL, simplified sharing, and Identity and Access Management (IAM) for greater security, BigQuery further democratizes cloud data analytics.
cloud.google.com/blog/big-data/2016/09/bigquery-introducing-powerful-new-enterprise-data-warehousing-features
Machine learning built for business.
The ability to warehouse data means gaining insight from multiple data streams is all the more challenging—and critical. Now in public beta, Cloud Machine Learning is now fully integrated with BigQuery, Cloud Dataflow, and Cloud Storage, to let you quickly create, train, and add machine learning to your operations and applications. Learn how to get started with our new certification program and professional services for machine learning.
cloud.google.com/blog/big-data/2016/09/google-cloud-machine-learning-now-open-to-all-with-new-professional-services-and-education-programs

As the core of any cloud workload, compute should be affordable, fast and scalable. That's why in October, we lowered the price of Compute Engine by 10% for all instance types in every region. On November 4th at Google Cloud Platform Live, we announced additional price reductions for networking and storage products.

Networking
As more of our customers manage traffic to and from instances and the Internet, we've lowered egress prices as much as 47% across Asia-Pacific to match prices in North America, Europe, the Middle East and Africa. Due to their particular costs, egress to China and Australia will be priced separately from the rest of Asia-Pacific, with China rates increasing March 1st, 2015. Developers can lower egress costs further by peering with Google via Google Cloud Interconnect at preferential rates.

Persistent Disk
The price for SSD Persistent Disk has dropped by 48% to $0.17 per GB/Month, still with no separate fee for I/O operations. Storage for Persistent Disk snapshots is now the same price as standard Google Cloud Storage: $0.026 per GB/Month, a 79% reduction. Beginning March 1st, 2015, restores of Persistent Disk snapshots will be subject to the same egress fees as other Compute Engine network traffic.

Cloud SQL
We've dropped the price of our biggestCloud SQL instances — D4, D8, D16, and D32 — by 25%. This applies to both Packaged and Per-Use billing plans.

BigQuery
The price of BigQuery storage has dropped by 23% to $0.02 per GB/Month, the same price as Google Cloud Storage's popular Durable Reduced Availability (DRA) storage class.

With some of the lowest prices in the industry, Compute Engine helps developers gain the performance and consistency of Google's worldwide network. Now it's more affordable than ever to deploy large clusters, handle giant data sets and work more efficiently in the cloud.

We know that developers don’t want to spend time managing their billing information so we’ve added a couple of useful features to make managing your billing easier and less time consuming.

First, you’re now able to consolidate multiple projects and pay for them under a single billing account. This means you no longer have to manage and maintain billing information in multiple places. It also lets you quickly and easily set up new projects without having to re-enter your billing details.



Second, we know that sometimes you want another person (e.g. your accountant) to have access to your billing information but don’t want to grant them ownership permissions on your projects. You can now invite additional billing administrators from the Developers Console who can view and manage your billing details but are not automatically project owners.



With these two changes, you can spend time less on billing and more time doing the work you care about.

-Posted by Dan Stokeley, Product Manager