Business-ready cloud AI platform from Oracle and Nvidia would include AI chatbots, recommender engine systems, and computer vision

Source: NVIDIA

Oracle CloudWorld 2022 unleashes major cloud, AI efficiencies for enterprise

Oracle unveiled a host of cloud, data, and innovation updates at its CloudWorld 2022 expo. Among the notable highlights was the reveal of Oracle Alloy, a new cloud infrastructure platform with which Oracle’s partners can introduce new cloud services to their own customers.

There was also the introduction of MySQL HeatWave Lakehouse, which helps customers process and query hundreds of terabytes of data and combines transaction processing with data-driven analytics and intelligent automation within a single MySQL database.

Not to be left out were the collection of Oracle Cloud, Database, and partner ecosystem launches and enhancements, such as the digital transformation pact with HSBC to migrate business-critical databases and systems to Oracle Exadata Cloud@Customer, a cloud platform delivered as a managed infrastructure service within the financial services provider’s own data centers – powering cloud automation to manage HSBC’s legacy applications, while providing the robust facilities to more easily scale services according to customer and market demands.

Not to be outdone, two AI-centric platforms were also revealed, both running massive, scale-out NVIDIA H100-based data centre systems, which promise better energy efficiency and improved performance for running large AI models. Built collaboratively with Nvidia, social media juggernaut Meta revealed Grand Teton, an AI platform that builds on the company’s previous generation Zion EX platform with more memory, more network bandwidth, and more compute capacity.

Grand Teton leverages Nvidia’s H100 Tensor Core GPUs to train and process AI models at data center scales, providing greater computing capacities to enable the AI to grow rapidly in size and capabilities. The announcement came not long after Nvidia’s partnership with Oracle itself was declared to be expanded, by harnessing tens of thousands of Nvidia H100 GPUs to power Oracle Cloud Infrastructure – bringing a high-intensity, AI-enabled performance platform for enterprises.

According to Nvidia founder-CEO Jensen Huang and Oracle CEO Safra Catz, the growing partnership between the two digital leaders will make AI-centric cloud infrastructure more accessible and encourage innovative collaboration in key sectors, including smart manufacturing and telecommunications to the healthcare and financial services industries.

Thanks to the Nvidia Hopper AI architecture and the powerful new Transformer Engine, H100 will allow enterprises to significantly reduce the cost of delivering AI applications, while providing the reliable high level of performance that business AI use cases require. With the Oracle partnership, AI workflows harnessing the Nvidia AI Enterprise software suite can be developed and deployed specifically on Oracle Cloud Infrastructure.

In fact with the full Nvidia AI stack being introduced to Oracle Cloud Infrastructure, data processing for end-to-end AI workflows can be greatly accelerated by leveraging NVIDIA Accelerated Apache Spark in OCI Data Flow, with alterations easily made with no-code– optimizing data processing resources, and managing unnecessary additional costs that used to be incurred for the enterprise customer.

Business-ready cloud AI platform from Oracle and Nvidia would include AI chatbots, recommender engine systems, and computer vision


The additional processing power from the added-on GPU clusters (with up to 512 Nvidia GPUs per cluster) will be necessary to handle the high-performance AI file systems on Oracle Cloud Infrastructure such as WEKA, Lustre, GlusterFS or IBM Spectrum Scale – producing up to 2 terabytes of throughput per second, and capable of processing a variety of preconfigured or cost-optimized requirements.

With the Nvidia AI Enterprise suite enabling key abilities, including cloud-native management and AI and accelerated data science development tools on top of Oracle Cloud Infrastructure, ready-made AI frameworks and applications equipped with deployment capabilities for enterprises will be made available.

The range of business-ready AI applications would include conversational AI chatbots, recommender engine systems, and computer vision capabilities. And with a broad, certified partner ecosystem to integrate with, both development times and complexities can be dramatically cut down – always a major concern for companies. Additionally, the efficiency savings can then translate towards quicker turnaround on actual usage and data insights on performance that can be studied, and improved upon.

Along with the Oracle Cloud Infrastructure Data Flow layer accelerated with Apache Spark for more rapid data processing, Nvidia GPUs will also speed up TensorFlow acceleration for Spark natural language processing, machine learning and deep learning applications. Over those, a dedicated AI framework for healthcare and life sciences, Nvidia Clara on Oracle Cloud Infrastructure, has also been developed to advance AI research into medical devices, genomics analyses, medical imaging advancements, and advanced drug discovery studies.

“To drive long-term success in today’s business environment, organizations need answers and insight faster than ever,” commented Oracle CEO Safra Catz. “Our expanded alliance with NVIDIA will deliver the best of both companies’ expertise to help customers across industries — from healthcare and manufacturing to telecommunications and financial services — overcome the multitude of challenges they face.”

“Accelerated computing and AI are key to tackling rising costs in every aspect of operating businesses,” added Jensen Huang, the CEO and founder of Nvidia. “Enterprises are increasingly turning to cloud-first AI strategies that enable fast development and scalable deployment. Our partnership with Oracle will put NVIDIA AI within easy reach for thousands of companies.”