* total quarterly revenue was $21.5 billion, up 21% yearoveryear, and up 6% sequentially [Annual numbers would be about 4.2 times bigger]
* U.S. revenue was up 25% yearoveryear to $10 billion. U.K. revenue was up 14% yearoveryear to $1.9 billion
* Rest of world revenue was up 19% versus last year, to $9.6 billion.
* GAAP other cost of revenues were $4.2 billion. NonGAAP other cost of revenues was $3.9 billion, up 28% yearoveryear, primarily driven by Googlerelated expenses, specifically, costs associated with operating our data centers, including depreciation; content acquisition costs, primarily for YouTube; and hardware costs.
* GAAP operating expenses were $7.4 billion in the quarter. NonGAAP operating expenses were $6.2 billion or 29% of revenue, up 15% yearoveryear, and up 2% versus Q1
* NonGAAP operating income was $7.5 billion, up 25% versus last year. The operating margin was 35%.
* Headcount at the end of the quarter was 66,575, up 2,460 from last quarter. The vast majority of new hires continue to be engineers and product managers to support growth in priority areas such as Cloud and Apps
Google-ABC continues to invest in making search smarter and more useful.
In the U.S., they recently introduced a voter tool
A new keyboard for iOS called Gboard is off to a great start
At I/O, they shared a vision around the Google Assistant.
Google Assistant will be in products like our new messaging app Allo and our voiceactivated device, Google Home
Machine Learning is the key to Google's Future
Machine learning is the engine that will drive Google's future
More than 100 teams are currently using machine learning at Google, from Street View to Gmail to Voice Search and more. For example, in Search, we use a ranking signal called RankBrain, which relies on deep learning to improve results. It's already enhancing the search experience in 40 languages. And based on user testing, RankBrain can accurately guess which results users will favor with about 80% accuracy. Advances like this help us make our search results even more relevant.
Machine learning is also creating an impact in other ways. Just last week, we announced a test that applied DeepMind's machine learning to our own Google data centers, resulting in up to a 40% reduction in the energy we use for cooling.
They introduced two Cloud Machine Learning APIs for speech and natural language to help our enterprise customers convert audio to text and easily understand the structure and sentiment of the text in a variety of languages.
They introduced Tensor Processing Units, or TPUs, which can deliver an order of magnitude better optimized performance per watt for machine learning projects
A big part of making information available is making great content accessible, such as on Google Play, YouTube, and the Web. Our efforts to make the mobile Web better for everyone with Accelerated Mobile Pages, or AMP, has seen phenomenal global momentum. Google now has over 150 million AMP pages in their index, and with over 4 million new ones published every week from nearly 200 countries. They announced AMP for Ads, which helps advertisers build fast, simple, and beautiful ads and landing pages
Google for Everyone
They are working with Indian Railways and RailTel to bring highspeed wireless access to the entire Internet to millions of people who travel throughout India's top railway stations. There are already two million people logging in every month, and they are using as much as 15 times the data they would otherwise use in a full day on their cellular networks.
They are also working hard to ensure that their products work well for everyone, regardless of where they live. We have been expanding direct carrier billing in Google Play, with more than 100 partners in 40 countries. To give you a sense of the impact, in Indonesia alone, they have seen growth in monthly buyers quadruple in the last 18 months.