Oracle has announced a big collaboration with AMD to construct a large-scale computing cluster using 30,000 AMD MI355X GPUs. According to Oracle CTO Larry Ellison, the arrangement, announced during the company’s most recent earnings call, represents a multibillion-dollar investment.

The AMD MI355X GPUs, which are slated to be released in mid-2025, include cutting-edge features such as 288GB of HBM3E memory, an 8Tbps memory bandwidth, and a 1,100W power demand that requires liquid cooling. These GPUs, based on AMD’s CDNA 4 architecture and manufactured using TSMC’s 3nm process, are expected to compete directly with Nvidia’s future B100 and B200 models.

Oracle has not revealed the location of the new cluster, but it follows another significant development: a collaboration with OpenAI to deploy 64,000 Nvidia GB200 GPUs in a data center in Abilene, Texas. While Oracle has yet to conclude contracts with OpenAI for the Stargate project, Ellison expects them to be finished soon.

Oracle’s current initiative builds on its earlier partnership with AMD, which included the integration of AMD Instinct MI300X GPUs into its OCI Compute Supercluster, which supports up to 16,384 GPUs. Although Oracle continues to be a significant user of Nvidia’s AI processors, this purchase demonstrates the company’s dedication to diversifying its GPU portfolio and researching high-performance alternatives.

Oracle’s collaboration with AMD demonstrates its commitment to enhancing AI and high-performance computing capabilities. As the industry expects future breakthroughs, this collaboration might have far-reaching ramifications for cloud infrastructure and AI workloads.

As Oracle increases its expenditures in high-performance computing and AI infrastructure, it may be positioned itself to challenge Nvidia’s supremacy in the GPU compute service market. Oracle’s 23ai database is already designed to accept native vector data formats, allowing for smooth integration with corporate information—a critical feature for enterprises seeking to harness AI-driven insights.

Oracle’s long-term ambition for hardware-software integration has proven accurate. The company’s $7.4 billion acquisition of Sun Microsystems in 2010, formerly regarded as a risky move, has paid off handsomely. Since then, Oracle has released 13 generations of Exadata database machines, effectively making them into database supercomputers. Oracle’s latest Exadata X11M, which launched in January this year, demonstrates the company’s ongoing commitment to enterprise computing innovation.

Oracle is well-positioned to grow its cloud business, thanks to its extensive experience in hardware-software co-design and a large installed base across industry, government, academic, and research institutions. In less than a decade, Microsoft used its huge Windows Server and SQL Server footprint to become the second-largest cloud provider; Oracle’s entrenched database supremacy might fuel a similar trajectory in the cloud industry.

Oracle’s ability to provide customers with Oracle-engineered infrastructure within AWS, Microsoft Azure, and Google Cloud is a critical distinction. This flexibility is especially appealing to businesses that choose Oracle’s proprietary technologies over cloud providers’ hardware solutions. Oracle’s supremacy in high-end relational databases for enterprises has been maintained by retaining this unique offering.

Oracle’s attraction is further enhanced by the fact that it offers customized, dedicated cloud regions to customers that want more control over their cloud environment. Its cloud services are compatible with on-premises systems, offering seamless management, uniform billing, and a consistent user experience. This method, inherited from Sun Microsystems’ designed systems, enables Oracle to provide customized cloud installations while harnessing insights from its extensive customer base.

Oracle is establishing itself as a serious player in the high-performance computing market with increased expenditures in AI infrastructure, database innovation, and cloud flexibility. As demand for AI-driven enterprise solutions grows, Oracle’s strategy has the potential to fundamentally transform the industry’s competitive dynamics.