Google Cloud Platform (GCP) Product Overview
Before Getting Started…
This page covers some of the core services Google Cloud Platform offers.
Before getting started, register for a GCP account. Each new user gets $300 in account credits to spend in the first 12 months.
GCP offers virtual machines for general use; the most commonly used service is Compute Engine.
GCP also has specialized offerings for specific types of workloads. This includes services which have dedicated GPUs and can perform computationally intensive tasks. A complete look at the VM product suite is available here.
Cloud Functions are used to run custom code when asked, without having to provision, maintain, and pay for the infrastructure it runs on. It is a close analogue to AWS Lambda.
GCP has many database solutions for different needs, and the most common ones are summarized below:
- Cloud SQL is a fully managed relational database solution.
- Firestore is a NoSQL key-value database solution which was created with mobile, web, and IoT applications in mind.
- Bigtable is a NoSQL key-value database solution which is ideal for large (petabyte-scale) datasets, data analytics, and machine learning.
For a full list of database offerings in GCP, see the GCP Databases Overview page.
Cloud Storage is used to store, back up, and retrieve files, videos, logs, and anything else you can imagine. It is a close analogue to AWS Simple Storage Service (S3).
You can serve a static website via Cloud Storage in a similar fashion to S3; check out the GCP static website tutorial for more information.
Dedicated file storage for one or more virtual machines is provided through Persistent disk.
Firebase is the seminal GCP service for creating, deploying, and managing mobile applications. It includes features like:
- Native integration with GCP services like Firestore, Cloud Functions, and Cloud Storage
- Testing for phsyical Android and iOS devices with Test Lab
- App crash reporting and analysis
App Engine is a fully-managed, highly scalable service which deploys an application you write (in any common language) to the web.
App Engine is particularly useful for writing web applications and supporting an API-based backend for mobile applications. App Engine integrates tightly with Firebase for seamless developer and user experiences.
For more information about setting up App Engine, check out the how-to guides in the App Engine documentation.
Security and Permissions
Read about GCP permissions management with IAM on our documentation page.
Power User Products
GCP has a product suite for common needs for large businesses and enterprise shops. The most notable product families and their core services are listed here for reference.
GCP offers general-purpose and specialized services for machine learning applications.
- AutoML is a general-purpose product for training and deploying machine learning models in GCP.
- AI building blocks supports easy integration of machine learning into other GCP applications and projects.
- Vision AI and Video AI facilitate model training and deployment for computer vision problems.
- Natural Language provides core capabilities of natural language processing in the cloud.
- TTS, STT, and Dialogflow enable human-computer interaction for a variety of platforms and applications.
View the full suite of machine learning offerings on the GCP website.
For large-scale data analytics, GCP offers several key services which are purportedly more cost-effective than many competing options.
- BigQuery enables low-latency, cost effective data analysis for large (petabyte-scale) datasets.
- Dataflow provides batch processing functionality for dataset queries, while abstracting away server management from the scientists and engineers who are writing and running queries.
- Dataproc creates open-source data analysis solutions, including native support of Apache Spark, Apache Hadoop, and other major platforms.