Product
Ocient Favicon
The Ocient Hyperscale Data Warehouse

To deliver next-generation data analytics, Ocient completely reimagined data warehouse design to deliver real-time analysis of complex, hyperscale datasets.

Learn More
Pricing Icon
Pricing

Ocient is uniquely designed for maximum performance and flexibility with always-on analytics, maximizing your hardware, cloud, or data warehouse as a service spend. You get predictable, lower costs (and absolutely zero headaches).

See How
Solutions
Customer Solutions and Workload Services Icon
Customer Solutions and Workload Services

Ocient offers the only solutions development approach that enables customers to try a production-ready solution tailored to their business requirements before investing capital and resources.

Explore
Management Services Icon
Management Services

Tap into the deep experience of the Ocient Management Services team to set up, manage, and monitor your Ocient solution.

Learn More
Company
Ocient Favicon
About Ocient

In 2016 our team of industry veterans began building a hyperscale data warehouse to tackle large, complex workloads.

Learn More
Published January 29, 2025

Ocient Collaborates with AMD to increase compute efficiency and decrease customers’ TCO

Jenna Boller, Ocient teamBy Jenna Boller, VP of Marketing

Today, we announced a collaboration with AMD that has been a long time in the making and that has resulted in remarkable gains for our company and our customers – a 3.5x increase in processing power, more than 2x increase in memory throughput, and lower energy consumption for compute-intensive data processing, analytics, and predictive AI workloads delivered by the Ocient Hyperscale Data Warehouse™ running on 4th Gen AMD EPYC processors.

Read the press release: “Ocient Announces Collaboration with AMD to Deliver Enhanced Power Efficiency and Performance for Data and AI Workloads”

While this is an exciting step along Ocient’s journey in delivering industry-leading efficient and unified data analytics solutions for our enterprise and public sector customers, it’s one of many past, present, and future moves we’re making to lower the cost, footprint, and energy utilization for always-on, compute-intensive data analytics and AI workloads.

Optimizing for Compute Efficiency with Leading Industry Partners

Since our founding, Ocient has optimized our software architecture and design patterns to run on modern, industry-standard hardware, ensuring the latest developments in hardware innovation are actualized in customer deployments. We collaborate with industry leaders to align our product roadmap and feature development with advances and breakthroughs across the product roadmap. With each new foundational leap, Ocient remains on the cutting edge, delivering industry-leading performance at scale while lowering the cost of providing complex data processing and analytics capabilities to our customers.

As a result, our customers have been able to drastically reduce the system footprint of their data and AI workloads. Ocient’s continued focus and innovation around maximizing compute per Watt not only reduces the carbon intensity of always-on data processing and analytics, but also enables our customers to contribute to key sustainability goals without impacting performance.

What’s Behind Ocient’s Collaboration with AMD

Together, Ocient and AMD deliver exceptional performance optimized for a wide range of analytical and AI workloads. Thanks to the 4th Gen AMD EPYC processors, Ocient customers can accelerate these workloads with high-per-core performance in order to:

  • Pursue new product innovation, intelligence, and revenue streams that rely on larger, more complex data sources and datastores
  • Modernize their data center footprint in a way that maximizes hardware efficiency
  • Reduce energy consumption for always-on, compute-intensive data processing and analytical workloads
  • Minimize TCO while optimizing results for their business

In many of our joint customer scenarios – and particularly for on premises deployments – data center space is at a premium. Energy demand is on track to quickly outpace supply, yet the need for high performance computing solutions only continues to grow.

By co-developing next-generation solutions with key industry partners like AMD, Ocient is expanding “the art of the possible” and fueling innovation for industries and use cases that were previously unable to harness high volumes of data in a performant and cost-effective way.

 

Why Core Density Matters

One of the key benefits of this collaboration includes Ocient’s ability to leverage the incredible core density of the 4th Gen AMD EPYC processors. Core density is crucial for accelerating analytics on large-scale datasets because it directly correlates to gains in performance, efficiency, and scalability. In analytical environments, increased core density unlocks higher levels of concurrency with fewer bottlenecks and massively parallel processing power.

For AI-driven workloads, higher core density translates to better predictive insights, faster model training, and the ability to handle growing data volumes efficiently.

In either scenario, Ocient’s Compute Adjacent Storage Architecture® (CASA) with AMD’s EPYC processors direct attached to high-capacity NVMe drives increases data throughput, lowers latency for large-scale computations and real-time AI inference, and enables faster and better decision-making.

Ocient reduces system sizing and energy consumption for compute intensive data
In addition to pioneering Compute Adjacent Storage Hardware (CASA), Zero Copy Reliability, Megalane, and more, Ocient leverages large capacity NVMe drives and power-optimized CPUs to deliver a drastic reduction in system cost, sizing, and energy consumption

Through our collaboration with AMD, Ocient delivers more cores per dollar for always-on, compute-intensive workloads that can otherwise be cost prohibitive to scale in the cloud or on premises with a less efficient data platform solution. By incorporating higher core-count CPUs into the underlying foundation and working closely with partners like Solidigm to optimize for large capacity NVMe solid state drives, Ocient decreases the cost, system footprint, and energy consumption for high performance data processing, analytics, and predictive AI by up to 90%.

 

How Ocient Maximizes Compute Per Watt

In addition to showcasing our gains in processing power and memory throughput, we are excited to collaborate with AMD to optimize Ocient’s CPU resource utilization for compute-intensive tasks that require high memory bandwidth and processing power. By optimizing for efficiency across high core count processors, Ocient optimizes power consumption per computational unit and improves sustainability via the amount of compute delivered per watt on premises and in the cloud.

As a result, our customers have been able to drastically reduce the system footprint of their data and AI workloads. Ocient’s continued focus and innovation around maximizing compute per Watt not only reduces the carbon intensity of always-on data processing and analytics, but also enables our customers to contribute to key sustainability goals without impacting performance.

What’s Next for Ocient and AMD

With all of our hardware partners, Ocient is on a never-ending mission to maximize price-performance across the entire hardware-software stack. As we continue co-designing hardware-enabled software solutions with partners like AMD, we will continue to launch new use case solutions enabled by next-generation breakthroughs in performance, scale, and energy efficiency. In the meantime, if you’re ready to see Ocient in action, reach out to our team and get in touch.