(MENAFN- The Arabian Post)

Global enterprise cloud provider Oracle Corporation is pushing its multicloud and artificial intelligence strategy into a new phase, deploying ultra-high-density GPU clusters, open standards and unified data architecture to support large-scale model training and agentic AI workloads. At the heart of the initiative is Nathan Thomas, Vice President of Multicloud and AI Strategic Initiatives, who described the engineering-driven enhancements unveiled at the company's AI world event in Las Vegas this week.

Thomas said the foundation for enterprise AI must begin with“extremely strong base layers” of infrastructure, adding that several of Oracle's latest announcements focus precisely on that. He pointed to innovations such as the OCI Zettascale 10 platform that supports up to 131,000 GPUs in a single cluster, paired with the Acceleron network fabric and Helios rack architecture. These enable tightly coupled compute and inference workloads, reducing inter-node latency and supporting distributed training at hyperscale.

The strategy around multicloud is crucial. Oracle is enabling its database and AI services to run across major cloud platforms-Amazon Web Services, Microsoft and Google LLC-via dedicated interconnects and consistent hardware footprints. Thomas explained that enterprises can retain their Oracle database architecture while deploying AI pipelines on third-party clouds, maintaining governance and compliance. He also noted that Oracle will expand from 38 regions to 72 regions globally to execute this vision.

On the infrastructure front, Oracle is adopting both Nvidia and AMD GPU stack solutions. Thomas said that while Nvidia's GB200 and GB300 clusters remain central, AMD's MI355X and MI450 are gaining momentum under a heterogeneous-compute model. He emphasised that choice and flexibility matter to customers who run highly varied workloads across industries. The advanced Helios rack supports 72 GPUs per rack, liquid cooling and unified accelerator networking to boost density and reduce latency - a departure from traditional modular designs.

See also Unconventional Approaches to AI Prompt Engineering Boost Accuracy

The platforms being built are designed for agentic AI and large language models, not just standard cloud services. Thomas outlined how Oracle's AI Data Platform unifies structured and unstructured data across transactional systems, vector embeddings and document stores-an approach that Thomas said is key to enabling enterprise-grade AI rather than starting from scratch.

Markets such as India are also central to Oracle's global roadmap. Thomas cited more than 5,000 enterprise customers and over 500 partners in the country, noting that Oracle already operates two cloud regions there and intends to roll out AI training and inference-capable regions in the next year. India's nine product-development centres contribute significantly to Oracle's engineering operations. Alongside geographic expansion, Oracle emphasises sustainability. Thomas said capacity builds are demand-driven, cooling and power partnerships are localised and high-density deployment allows better energy efficiency and resource utilisation.

Competition remains intense, with other hyperscalers and AI-infra firms vying for dominance in large-scale AI compute. But Oracle is betting that a full stack-from high-density infrastructure to unified data to multicloud flexibility-will appeal to enterprise customers who face complex workloads, regulatory pressures and hybrid architectures.

Analysts note that while peak performance claims will need independent verification, the strategic direction suggests Oracle is positioning itself not just as a cloud provider, but as an infrastructure partner for enterprises scaling AI across clouds.

For customers, the implications are twofold: one, access to increased GPU scale, network performance and data-platform integration; two, greater freedom to distribute workloads across clouds without sacrificing the underlying database architecture. That dual proposition may resonate in regulated industries or international enterprises with complex IT estates.

See also Dubai Welcomes DoBeGo - New Hub for Car Purchase, Leasing, Rental, and Finance

Notice an issue? Arabian Post strives to deliver the most accurate and reliable information to its readers. If you believe you have identified an error or inconsistency in this article, please don't hesitate to contact our editorial team at editor[at]thearabianpost[dot]com. We are committed to promptly addressing any concerns and ensuring the highest level of journalistic integrity.

MENAFN02112025000152002308ID1110281912