PRIVATE ATTACHED STORAGE
Private Attached Storage for Data-Intensive Workloads
GPCN provides SSD and NVMe storage options to support private compute, enterprise applications, analytics, AI and recovery workloads.
THE DATA LAYER
Storage Is Part of the Workload
Storage decisions go beyond capacity. The right architecture must support the performance, availability and economics of the workload it serves.
GPCN integrates private storage with dedicated compute across a global network, giving organizations greater control over performance, data placement, scalability and movement.
Performance
Dedicated SSD and NVMe storage delivers predictable performance alongside the CPU and GPU resources supporting the workload.
Persistence
Keep data close to the applications and compute that depend on it, with private storage designed around the workload's performance and retention requirements.
Growth
Scale storage alongside CPU, GPU and memory as requirements change—without the capital investment and provisioning cycles of owned infrastructure.
Data Movement
Move data across the GPCN network without GPCN egress fees, helping reduce the cost and friction associated with data-intensive workloads.
Recovery
Build recovery capacity into the infrastructure strategy, with private compute and storage available across geographically distributed GPCN locations.
Location
Place infrastructure and data where the business requires it, leveraging GPCN's global availability zones to address performance, proximity and data residency requirements.
STORAGE OPTIONS
Match Storage to the Requirement
SSD
Persistent SSD capacity for common enterprise application, VM, backup and data workloads.
NVMe
Higher-performance persistent storage for workloads with more demanding I/O and throughput requirements.
- Included Compute StorageStandard GPCN CPU profiles include SSD capacity aligned to the VM configuration.
- Persistent Block VolumesAdditional persistent volumes can be provisioned as data and application requirements grow.
AI & DATA
Keep AI Storage and Compute in the Same Infrastructure Conversation
Production AI creates sustained storage requirements.
Training datasets, embeddings, vector stores, checkpoints, model artifacts and inference outputs can move repeatedly between storage, CPU and GPU resources.
GPCN allows persistent storage, private GPU and supporting CPU resources to be planned as one infrastructure design.
DATA MOVEMENT
Understand the Cost of Moving the Data
Large data volumes can make network-transfer economics an important part of storage architecture.
GPCN does not charge egress fees for data leaving or moving within GPCN.
Where practical, locating compute and storage closer to one another can help reduce unnecessary movement while also supporting latency, residency and operational requirements.
BUILD AROUND THE DATA
What Does Your Workload Need From Storage?
Tell us the capacity, performance, location and recovery requirements. We’ll help evaluate the complete workload.