At Google Cloud Next '23, a highly anticipated tech event, Google Cloud made significant announcements that are set to revolutionize the industry. The company unveiled new ways to harness the power of generative AI and leading cloud technologies, providing businesses, governments, and users with enhanced capabilities. From AI-optimized infrastructure to powerful tools like Vertex AI and Duet AI, Google Cloud is empowering organizations to leverage the full potential of AI and drive innovation.
AI-Optimized Infrastructure: The Backbone of Revolutionary Gen AI
Underpinning the advancements in generative AI is Google Cloud’s state-of-the-art AI-optimized infrastructure. With more than 25 years of investment in data centers and networks, Google Cloud boasts a global network of 38 cloud regions. The company has set an ambitious goal to operate entirely on carbon-free energy by 2030, demonstrating its commitment to sustainability.
Google Cloud’s infrastructure serves as a leading choice for training and serving gen AI models. More than 70% of gen AI unicorns, including industry giants like AI21, Cohere, and Runway, are already Google Cloud customers. Additionally, over half of all funded gen AI startups, such as Copy.ai and Fiddler AI, have partnered with Google Cloud for their infrastructure needs.
To further assist customers in their AI journey, Google Cloud announced key infrastructure advancements
Cloud TPU v5e: This highly cost-efficient and scalable AI accelerator is designed to handle large-scale AI training and inference workloads. Offering up to a 2x improvement in training performance per dollar and up to a 2.5x improvement in inference performance per dollar compared to its predecessor, Cloud TPU v4, it empowers businesses of all sizes to drive AI innovation.
A3 VMs with NVIDIA H100 GPU: These powerful virtual machines, powered by NVIDIA’s H100 GPU, provide organizations with three times better training performance than the previous A2 generation. With high-performance networking and other cutting-edge features, they enable businesses to tackle the most demanding gen AI and large language model (LLM) applications.
GKE Enterprise: Designed for mission-critical AI/ML workloads, GKE Enterprise facilitates multi-cluster horizontal scaling. By leveraging autoscaling, workload orchestration, and automatic upgrades, businesses can achieve significant productivity gains of 45% and reduce software deployment times by over 70%. The availability of Cloud TPU v5e further enhances GKE’s capabilities.
Cross-Cloud Network: This global networking platform helps customers securely connect applications across multiple clouds. By reducing network latency by up to 35%, it enables seamless access to Google services from any cloud environment. ML-powered security ensures zero trust, safeguarding critical data.
Google Distributed Cloud (GDC): GDC caters to organizations that require edge computing or on-premises workloads. With Vertex AI integrations and a new managed offering of AlloyDB Omni on GDC Hosted, businesses can harness AI capabilities closer to their operations, unlocking new possibilities for innovation.
Enhancing the Vertex AI Platform:
Google Cloud’s Vertex AI is a comprehensive platform that enables customers to build, deploy, and scale machine learning (ML) models. During the event, Google highlighted the impressive growth of gen AI customer projects, which increased by over 150 times from April to July 2023. Vertex AI offers access to more than 100 foundation models, optimized for various tasks such as text, chat, images, speech, and software code.
To further enhance the capabilities of Vertex AI, several new models and tooling have been introduced. PaLM 2 has been upgraded to support processing longer-form documents like research papers and books. Improvements have also been made to Imagen’s visual appeal, and support for new languages has been extended in Codey.
In terms of tuning, adapter tuning for PaLM 2 and Codey, along with Style Tuning for Imagen, have been made available. These tools allow enterprises to customize and improve the performance of their models with minimal effort.
Additionally, Vertex AI has announced the availability of new models such as Llama 2, Code Llama, Falcon LLM, and pre-announced Claude 2. These models offer enhanced functionality and performance for various AI applications.
The platform also introduces Vertex AI extensions, which enable developers to access, build, and manage extensions that provide real-time information, incorporate company data, and take actions on behalf of the user. This opens up new possibilities for gen AI applications that can seamlessly integrate with enterprise systems, such as CRM or email.
An enterprise grounding service has been introduced to enable customers to ground responses in their own enterprise data, ensuring more accurate responses. Furthermore, a digital watermarking technology called DeepMind SynthID has been integrated into Vertex AI, offering a scalable approach to creating and identifying AI-generated images responsibly.
Colab Enterprise, a managed service provided by Vertex AI, combines the ease-of-use of Google’s Colab notebooks with enterprise-level security and compliance capabilities. Data scientists can leverage this service to accelerate AI workflows, access the full range of Vertex AI platform capabilities, and integrate with BigQuery.
Privacy and data control are crucial factors in AI development, and Vertex AI prioritizes these aspects. The platform ensures full control and segregation of customer data, code, and IP, with zero data leakage. Customers can train and customize their models using private documents and data without exposing them to the foundation model. This level of control empowers organizations to protect their sensitive information and maintain data privacy.
With its comprehensive features, new models, and robust data control mechanisms, Vertex AI continues to push the boundaries of AI technology. It offers an extensive platform for building and deploying AI models while prioritizing data privacy and customization, making it a valuable tool for businesses in various industries.
Looking Ahead:
Google Cloud’s announcements at Next '23 reflect the company’s commitment to democratizing access to AI and driving innovation across industries. By providing AI-optimized infrastructure and powerful tools like Vertex AI and Duet AI, Google Cloud empowers organizations to leverage the full potential of generative AI. As businesses continue to adopt AI technologies, Google Cloud’s advancements will undoubtedly shape the future of AI-driven innovation.
Conclusion:
Google Cloud’s announcements at Next '23 showcased its commitment to advancing the field of generative AI. With AI-optimized infrastructure, cutting-edge tools like Vertex AI, and collaborative solutions like Duet AI, Google Cloud is empowering businesses to unlock the full potential of AI. These advancements will accelerate innovation, drive productivity gains, and enable organizations to deliver better products and services to their customers. As the industry continues to evolve, Google Cloud remains at the forefront, driving the future of AI-powered technologies.
