What is Google Distributed Cloud?

Along with the rise of the cloud trend, more and more enterprises hope to accelerate the speed of cloud system adoption. They aim to leverage the advantages of the cloud to facilitate easier and quicker innovation in their research and development, while also reducing technical risks. However, limitations stemming from industry specifics, regional regulations, data sovereignty requirements, local data processing needs, or the necessity to coordinate with other services often make the adoption of cloud systems for businesses more challenging and complex.

Google Distributed Cloud(GDC) is an integrated solution combining software and hardware to assist enterprises in achieving seamless multi-cloud, on-premises data, and edge environment management through a single platform. This allows them to focus on application development and business strategies without worrying about underlying infrastructure. Depending on various deployment scenarios, enterprises can run Google Distributed Cloud in the following four environments:

Google Edge Network: Leveraging Google’s global network with over 140 edge computing points.
Carrier Edge: Edge networks owned by telecommunications carriers (CSPs). These are suitable for applications and services with strict requirements for low latency and stability due to the provision of 5G/LTE services.
Customer Edge: Suitable for industries like retail and manufacturing that require on-site computational processing at the edge to ensure data processing efficiency and operational effectiveness.
Customer Data Center: Supporting customer-owned data centers and co-located facilities to meet stringent data security and privacy requirements.

Google Distributed Cloud aims to meet the unique needs of organizations that perform work at the edge or in data centers but still require cloud service functionalities, flexibility, and scalability. Through Google Distributed Cloud, users can not only migrate or modernize applications using Google Cloud services but also perform localized execution including data processing, database analytics, machine learning, and more. Additionally, to enhance development efficiency, Google Distributed Cloud offers support for third-party services. This allows customers to integrate services from third-party providers within their dedicated development environments, fostering innovative operational models and improving the customer experience.

Three Hosting Modes of Google Distributed Cloud, Cater to Various Application Scenarios in Enterprises

Google Distributed Cloud is built on Anthos infrastructure, integrating on-premises data processing, edge computing, and modernized edge deployments. It provides enterprises with a high-performance, highly secure, and privacy-compliant multi-environment application management platform. Catering to different application scenarios, Google Distributed Cloud currently offers three primary hosting modes, enabling enterprises to choose the most suitable service options based on their needs.

Google Distributed Cloud Edge

Google Distributed Cloud Edge is a fully managed product focused on bringing computing and data processing capabilities closer to the source of data generation. It’s the ideal choice for running local data processing, low-latency edge computing workloads, modernizing development environments, and deploying private 5G/LTE solutions within enterprises. By being closer to the data source environment, it enables real-time access to necessary analytics data, reducing data transmission latency and enhancing application responsiveness.

Google Distributed Cloud Hosted

Google Distributed Cloud Hosted is an independent cloud platform that enables users to achieve modernization through secure and reliable on-premises deployment. It allows managing infrastructure, services, APIs, or tools without the need for a constant connection to Google Cloud, utilizing the local control plane provided by Anthos. This platform offers several key advantages, including openness by leveraging open source and commercial hardware, integrating industry innovation, and an open ISV ecosystem. By providing a consistent application experience, Google Distributed Cloud Hosted harnesses its experience, validated large-scale best practices, and technology from Google’s core services to meet diverse user data analytics needs, ensuring secure, intelligent, and modern cloud solutions.

Google Distributed Cloud Virtual

Google Distributed Cloud Virtual integrates Google software services with enterprise hardware, allowing businesses to concurrently utilize their own systems to run services while managing connections through Google’s control plane. Built on Anthos on-prem, Google Distributed Cloud Virtual can integrate virtual machines including VMware’s vSphere and Anthos bare metal. Hence, it is well-suited for enterprises adopting virtualized environments or planning to gradually migrate workloads to the cloud.

Optimizing with Integrated AI and Data Analytics for Greater Flexibility

At this year’s Google Next ’23 conference, Google announced the addition of three new solutions for Google Distributed Cloud services, further meeting the unique needs of enterprises in edge computing or data analytics workloads.

Vertex AI：
Google Distributed Cloud Hosted’s Vertex AI offers pre-trained models for voice, translation, optical character recognition (OCR), and Workbench. Integrating Vertex Prediction and Vertex Pipelines into Google Distributed Cloud Hosted assists enterprises in customizing automated service models and efficiently managing the execution of machine learning operations at scale. Additionally, Google Distributed Cloud Hosted introduces an API translation service, supporting document translation and format conversion for Microsoft Office and PDF files, thereby saving valuable time and resources for developers and their users.
AlloyDB Omni：
Google Distributed Cloud Hosted現在可於預覽版中支援AlloyDB Omni做為新的託管資料庫引擎。AlloyDB Omni除承襲了包含效能高、可完全相容於PostgreSQL、可簡化機器學習的自動管理功能…等AlloyDB既有優點外，在工作負載上AlloyDB更有著高於標準PostgreSQL2倍以上的處理效能，並能提供達100倍以上的分析查詢速度。AlloyDB Omni現也可支援pgvector的擴展，幫助用戶可以儲存、索引和查詢向量嵌入。Google Distributed Cloud Hosted now supports AlloyDB Omni as a new hosted database engine in the preview version. AlloyDB Omni inherits the existing advantages of AlloyDB, including high performance, full compatibility with PostgreSQL, and simplified machine learning automated management functions. Additionally, AlloyDB Omni exhibits over twice the processing performance of standard PostgreSQL for workloads and provides over 100 times faster analytical query speeds. AlloyDB Omni now also supports the pgvector extension, enabling users to store, index, and query vector embeddings.
Dataproc Spark：
許多企業使用Apache Spark等開源技術服務進行資料處理與分析。透過Google Cloud 用於運行開源資料工具的託管服務Dataproc，有助於客戶可以將Spark 的 TCO 降低百分之五十四。Google Distributed Cloud Edge現將可支援 Dataproc，以便用戶可以實現更好的節省、安全性和可管理性，同時支援資料駐留和監管要求。此服務預計將於 2023 年第四季推出預覽版。Many enterprises use open-source technology services like Apache Spark for data processing and analysis. Utilizing Google Cloud’s hosted service Dataproc for running open-source data tools helps customers reduce the TCO (Total Cost of Ownership) for Spark by up to 54%. Google Distributed Cloud Edge will now support Dataproc, allowing users to achieve better savings, security, and manageability while supporting data residency and compliance requirements. This service is expected to launch its preview version in the fourth quarter of 2023.

The latest version of Google Distributed Cloud offers various scales and configurations to meet deployment needs for independence and control. Primarily in the infrastructure, Google Distributed Cloud boasts higher performance, flexibility, and modular expansion options. These hardware features include fourth-generation Intel® Xeon® Scalable processors and high-performance network structures capable of up to 400 Gbps. Additionally, on the hosted hardware front, Google Cloud incorporates DDR5 high-bandwidth memory, Gen5 PCIe I/O performance, 400G/100G network connections, and dedicated GPUs designed for AI workloads, featuring Nvidia A100 chips that support airborne AI and data processing applications. Through these optimizations, users can better handle unstable network connections and, with simplified deployment models, can deploy fully managed applications across thousands of locations and scale management from a single central location.

Apart from the aforementioned feature updates, at the Next ’23 conference, Google announced the launch of a brand-new version of GKE – GKE Enterprise. This release aims to enhance cross-team collaboration, securely run critical business workloads, and reduce total operational costs through a fully integrated solution. As Google Cloud’s latest managed solution, GKE Enterprise will be included across all Google Distributed Cloud products at no additional cost. Moreover, to enable developers to create offline applications more easily, users can now run Google Distributed Cloud directly on an open development platform, utilizing open-source tools and ISVs such as ElasticSearch, MongoDB Enterprise Advanced, and SAP on Google Distributed Cloud to accelerate application development.

Google Distributed Cloud facilitates cloud computing expansion across regions, enabling resource sharing among multiple network systems. This allows each system host to focus solely on handling a single task, resulting in quicker completion of the overall task. As a Google Cloud Premier Partner, MacroTech assists various renowned enterprises in constructing cloud infrastructure and data application solutions across media, e-commerce, retail, information technology, and gaming industries. Feel free to fill out the contact form to let MacroTech’s dedicated consultants help plan the most suitable comprehensive cloud architecture solution for you!

This article is a rewrite from the official Google Cloud blog: source 1, source 2.

What is Google Distributed Cloud? Understanding Google Cloud’s Distributed Cloud Services