DigitalOcean Elevates its Agentic Inference Cloud with GPU Droplets powered by AMD Instinct™ MI350X GPUs

Key Terms

gpu technical

A GPU (graphics processing unit) is a specialized computer chip designed to handle many calculations at once, originally for rendering images and video but now widely used for tasks like artificial intelligence, data analysis and high-performance computing. Investors watch GPU demand and prices because strong sales often signal growth for chip makers and their customers, affect profit margins and capital spending, and can forecast wider trends in gaming, AI adoption and cloud services.

generative ai technical

Generative AI is a type of computer technology that can create new content, like text, images, or music, on its own. It’s important because it can produce realistic and useful material quickly, which could change how we create art, write stories, or even develop new products. Think of it as a smart robot that can invent and produce things almost like a human.

high-performance computing technical

A cluster of very powerful computers, special chips and fast networks designed to tackle huge, complex calculations far faster than a normal PC — like replacing a single delivery van with a synchronized fleet to move a city’s worth of packages. For investors, high-performance computing matters because it enables faster product development, more accurate simulations and data analysis, and new revenue streams for hardware, software and services, making firms that supply or use it potentially more competitive and scalable.

architecture technical

The overall design and organization of a system, product, or business — like the blueprint for how pieces fit together and communicate. For investors, architecture matters because it shapes a company’s ability to grow, control costs, adopt new technology, and manage risk; stronger, simpler designs often mean faster launches, lower maintenance bills, and better long-term competitiveness, while poor designs can lock in problems and expenses.

latency technical

Latency is the time delay between when information or an instruction is created and when it is received, processed, or acted on by a market system or data feed. For investors, that delay can alter the price you receive, cause missed trading opportunities, or increase execution risk — like sending a text to buy an item and the seller acting a few seconds later after the price has changed.

throughput technical

Throughput is the amount of stuff, like products or data, that a system can handle or move through in a certain period of time. For example, a factory’s throughput is how many items it produces each hour, and it matters because higher throughput usually means things are running efficiently and meeting demand quickly.

token technical

A token is a digital unit of value created on a blockchain that can represent money, access rights, ownership stakes, or unique items; think of it like a digital ticket or coupon that can be traded, used, or collected. Investors care because tokens can carry financial value, entitle holders to returns or governance rights, and their supply, legal status, and market demand affect price and risk in ways similar to stocks or commodities.

DigitalOcean and AMD collaborate to deliver new GPUs with lower latency and higher throughput for complex inference workloads

BROOMFIELD, Colo.--(BUSINESS WIRE)-- DigitalOcean (NYSE: DOCN), the Agentic Inference Cloud built for production AI at scale, today announced the availability of new, high-performance GPU Droplets powered by AMD Instinct™ MI350X GPUs. By integrating these GPUs into its Agentic Inference Cloud, DigitalOcean continues to deliver cost-efficient, high-performance solutions for leading AI-native companies like ACE Studio to scale their inference workloads. Next quarter, DigitalOcean will also deploy AMD Instinct™ MI355X GPUs, marking the addition of liquid-cooled racks to their offering and further expanding access to accelerators specifically designed for larger datasets and models on their inference cloud.

Optimizing production inference with AMD Instinct™ MI350X GPUs

AMD Instinct™ MI350X Series GPUs set a new standard for generative AI and high-performance computing (HPC). Built on the AMD CDNA™ 4 architecture, these GPUs deliver cutting-edge efficiency and performance for training massive AI models, high-speed inference, and complex HPC workloads including scientific simulations, data processing, and computational modeling. The capabilities of the GPUs allow DigitalOcean to optimize for compute bound prefill phase, while enabling high-performance inference at low latency and high token generation throughput. This provides the ability to load large models and larger context windows, leading to supporting a higher inference request density per GPU. Paired with DigitalOcean’s optimized inference platform, these feature enhancements of AMD Instinct™ MI350X GPUs offer lower latency and higher throughput.

“These results demonstrate that the DigitalOcean Agentic Inference Cloud isn't just about providing raw compute, but about delivering the operational efficiency, inference optimizations, and scale required for demanding AI builders,” said Vinay Kumar, Chief Product and Technology Officer at DigitalOcean. “The availability of the AMD Instinct™ MI350X GPUs, combined with DigitalOcean’s inference optimized platform offers our customers a boost in performance and the massive memory capacity needed to run the world’s most complex AI workloads while delivering compelling unit economics.”

Earlier this year, DigitalOcean announced that by optimizing AMD Instinct™ GPUs, they were able to deliver 2X production request throughput and a 50% reduction in inference costs for Character.AI, a leading entertainment platform with one of the most demanding product inference workloads. Similarly, customers like ACE Studio are building with AMD Instinct™ MI350X GPUs to power complex inference workloads while managing costs. “At ACE Studio, our mission is to build an AI-driven music workstation for the future of music creation,” said Sean Zhao, Co-Founder & CTO. “As we expand our footprint on DigitalOcean, the next-generation AMD Instinct™ MI350X architecture, supported by close collaboration on inference optimization with AMD and DigitalOcean, provides us a strong foundation to push performance and cost efficiency even further for our customers.”

"Our collaboration with DigitalOcean is rooted in a shared commitment to pairing leadership AI infrastructure with a platform designed to make large-scale AI applications more accessible to the world’s most ambitious developers and enterprises,” said Negin Oliver, Corporate Vice President of Business Development, Data Center GPU Business at AMD. “By bringing the AMD Instinct™ MI350 Series GPUs to DigitalOcean’s Agentic Inference Cloud, we are empowering startups and enterprises alike to deploy and scale next-generation AI workloads with confidence.”

This initiative builds upon previous collaboration between DigitalOcean and AMD, including the launch of the AMD Developer Cloud and the release of AMD Instinct™ MI300X and MI325X GPUs on DigitalOcean last year.

Enterprise performance with predictable cost-efficiency and simple operations at the forefront

In addition to offering the latest AMD GPUs, DigitalOcean maintains its commitment to transparency and simplicity, ensuring this powerful technology is easy to adopt for developers and emerging businesses:

• Cost-effective, predictable pricing: DigitalOcean offers transparent, usage based pricing with flexible contracts and no hidden fees.
• Simple setup: GPU Droplets can be provisioned and configured with security, storage, and networking requirements in just a few clicks, drastically simplifying deployment compared to complex cloud environments.
• Access to enterprise features: GPU Droplets offer enterprise-grade SLAs, observability features, and are HIPAA-eligible and SOC 2 compliant.

The new GPU Droplets powered by AMD Instinct™ MI350X are available in the Atlanta Region Datacenter.

To learn more about using AMD Instinct™ MI350X GPUs on DigitalOcean, please visit the DigitalOcean website.

About DigitalOcean

DigitalOcean is the Agentic Inference Cloud built for AI-native and Digital-native enterprises scaling production workloads. The platform combines production-ready GPU infrastructure with a full-stack cloud to deliver operational simplicity and predictable economics at scale. By integrating inference capabilities with core cloud services, DigitalOcean’s Agentic Inference Cloud enables customers to expand as they grow — driving durable, compounding usage over time. More than 640,000 customers trust DigitalOcean to power their cloud and AI infrastructure. To learn more, visit www.digitalocean.com

View source version on businesswire.com: https://www.businesswire.com/news/home/20260219844245/en/

Media Relations

Julie Wolf: press@digitalocean.com

Investor Relations

Melanie Strate: investors@digitalocean.com

Source: DigitalOcean