Google Cloud Next’23: Key AI announcements for the enterprise

Google Cloud Next’23: Key AI announcements for the enterprise
Photo Credit: 123rf.com

Google’s cloud division, Google Cloud, held its largest in-person event this week in San Francisco with an estimated 18,000 participants. The tech giant made a flurry of announcements around artificial intelligence (AI), and more specifically, generative AI at Google Cloud Next event.  

Google and Alphabet CEO Sundar Pichai said during a keynote on Tuesday that AI will touch every sector, every industry, every business function, and significantly change the way we live and work. “This isn’t just the future. We are already starting to experience the benefits right now. As a company, we’ve been preparing for this moment for some time," he added. 

Here are some key AI-related announcements the company made during the event that are aimed at enterprise users. 


Duet AI for Google Workspace 

A slew of new Duet AI features for Workspace were launched at the event. Google announced “Duet AI” as the branding for generative AI features in Workspace at I/O 2023 in May, which covers a range of generative AI tools for Google’s productivity apps, including Gmail, Drive, Slides, Docs, and more. The company now announced further Workspace AI integration in Google’s other core apps such as Meet and Chat.  

Aparna Pappu VP & GM for Google Workspace said, “Thousands of companies" and over one million people are already using Duet AI in Workspace.” She also revealed that Workspace now has over 10 million paying customers. Besides, there are over 300 new feature launches for Workspace in the last 12-months alone and these product investments appear to be paying off.  


In the coming weeks, Google said will add support for third-party apps to Chat, such as Zoho, Workday and Loom. 

Google also announced the integration of Duet AI for Mandiant Threat Intelligence, which will help CIO/CTOs gain natural language insights into security data, thereby enhancing productivity.   

Vertex AI gets better and better 


During the Cloud Next conference, Google announced updates to Vertex AI, the company’s generative AI platform that helps enterprises train their own AI and machine learning models. June Yang, vice president of cloud AI and industry solutions at Google Cloud, announced improvements to Vertex AI, as customers have asked for the ability to input larger amounts of content into PaLM, a foundation model under the Vertex AI platform. This led Google to increase its capacity from 4,000 tokens to 32,000 tokens. Also, Yang announced that PaLM is now available in Arabic, Chinese, Japanese, German, Spanish, and more, with a total of 38 languages for public use and 100 additional languages now in private preview. 

Further, Google Cloud has added Meta’s Llama 2 and Anthropic’s Claude 2 to Vertex AI’s model library. APIs of these models will be accessible through a new tool in its Cloud Platform named Model Garden. This makes Google Cloud the only Cloud provider to support first-party, open source and third-party models. That said,  

A3 virtual machine supercomputing with NVIDIA for AI training  


General availability of the A3 virtual machines running on NVIDIA H100 GPU as a GPU supercomputer will open in September, said Mark Lohmeyer, vice president general manager for compute and machine learning infrastructure at Google Cloud, during the event.  

The A3 supercomputers’ custom-made 200 Gbps virtual machine infrastructure has GPU-to-GPU data transfers, enabling it to bypass the CPU host. The GPU-to-GPU data transfers power AI training, tuning and scaling with up to 10 times more bandwidth than the previous generation, A2. The training will be three times faster, Lohmeyer said. 

Cross-Cloud Network to help businesses with multicloud networking 


Google announced an update to its Cloud Interconnect service. With the introduction of Cross-Cloud Interconnect, businesses can now securely connect any public cloud to Google Cloud through a high-performance network. The company further claims that using Cross-Cloud Network will reduce network latency by 35% and total cost of ownership by 40%, compared to connecting applications without routing the traffic over Google’s network. This advancement is expected to transform how CIO/CTOs handle the challenges of managing their multicloud workloads and dealing with the complexities of network connectivity and configuration. On the security front, Cross-Cloud Network is backed by Google’s ML-powered zero trust security suite. 

AI chip to handle faster training performance of LLMs  

Google Cloud has further announced advancements in its AI-optimised infrastructure, including fifth-generation TPUs and A3 VMs based on NVIDIA H100 GPUs dubbed Cloud TPU v5e. Traditional approaches to designing and constructing computing systems have now become inadequate with the growing demands of workloads like generative AI and large language models (LLMs), with CIO/CTOs realising the need for both cost-effective and scalable AI-optimised infrastructure. 


The cloud giant said Cloud TPU v5e is distinguished by its cost-efficiency, versatility, and scalability, the TPU aims to revolutionise medium- and large-scale training and inference. This iteration outpaces its predecessor, Cloud TPU v4, delivering up to 2.5x higher inference performance and up to 2x higher training performance per dollar for LLMs and generative AI models. 

Google brings its generative AI search experience to India 

Google has introduced generative artificial intelligence to its Search tool for users in India and Japan that will show text or visual results to prompts, including summaries. The new AI-powered search feature, also known as SGE (Search Generative Experience), was first launched only in the U.S and announced at this year's Google I/O Developer conference in May. SGE introduces a conversational mode to Google Search where you can ask Google questions about a topic and then have it return answers, similar to an AI chatbot. It will become available through Google's Search Labs in those markets, Google said, adding that users will have the choice to opt in for it. While Japanese users will be able to use the feature in their local languages, while it will be available in English and Hindi in India.  

"We are in an entirely new era of cloud, fuelled by generative AI," said Thomas Kurian, CEO, Google Cloud said at the event. "Our focus is on putting gen AI tools into the hands of everyone across the organization—from IT, to operations, to security, to the board room. As the industry's most open cloud, our goal is to help companies use AI and other cloud technologies to streamline their operations, increase productivity, and create entirely new lines of business,” he said.

Sign up for Newsletter

Select your Newsletter frequency