DigitalOcean Expands AI Cloud with AMD's Latest GPUs for Inference Workloads
Event summary
- DigitalOcean launched GPU Droplets powered by AMD Instinct™ MI350X GPUs on February 19, 2026.
- The new GPUs optimize compute-bound prefill phases and enable high-performance inference at low latency.
- DigitalOcean claims 2X production request throughput and 50% reduced inference costs for customers like Character.AI.
- Next quarter, DigitalOcean will deploy AMD Instinct™ MI355X GPUs with liquid-cooled racks.
- The new GPU Droplets are available in the Atlanta Region Datacenter.
The big picture
DigitalOcean's move to integrate AMD's latest GPUs into its Agentic Inference Cloud underscores the growing demand for cost-efficient, high-performance AI infrastructure. This strategic enhancement positions DigitalOcean to compete more effectively in the cloud computing market, particularly among AI-native enterprises requiring scalable and efficient inference solutions. The collaboration with AMD further solidifies DigitalOcean's commitment to providing enterprise-grade AI capabilities with predictable economics.
What we're watching
- Performance Scaling
- How the integration of AMD Instinct™ MI350X GPUs will affect DigitalOcean's ability to handle larger datasets and models.
- Market Differentiation
- Whether DigitalOcean can sustain its cost-efficiency claims amid increasing competition in the AI cloud market.
- Customer Adoption
- The pace at which AI-native companies adopt DigitalOcean's new GPU offerings for their inference workloads.
Related topics