Hammerspace and Azure
Unify, Automate and Accelerate Your Data in the Cloud
Unify Distributed, Siloed Data With a Global Namespace
that spans sites, cloud providers, and cloud regions
Migrate Data from On-Prem and Other Clouds to Azure
with Data-in-Place Assimilation and Data Orchestration
Run Demanding AI/HPC Workloads in the Cloud
with a standards-based, high-performance parallel file system
Simplify Hybrid-Cloud Infrastructure
By automating data movement between on-prem storage to cloud GPUs, without data copy
Streamline Cloud Data Pipelines and Workflows
with an end-to-end data platform that spans ingest through archive
Drive Compute
and Storage Consumption
with an agentless, cloud-native architecture that deploys on standard Azure Infrastructure
LLM and Gen AI Training
“What Hammerspace does is pure magic.”
– Principal Engineer, Meta
The Hammerspace Data Platform
Hammerspace helps Azure Cloud customers unlock the full value of their data by breaking down silos across on-premises, cloud providers, and cloud regions. Together with Azure, we make it easy to move data where it’s needed, speed up AI and HPC projects, and simplify hybrid-cloud operations. The result is faster time to value, lower infrastructure complexity, and greater consumption of Azure compute and storage.
Powering the
Most Demanding Data Workflows
AI Inferencing
Unifies unstructured data, prewarms datasets, and feeds GPUs at wire speed for low-latency inference.
AI Training
Combines Tier 0 and parallel file system to keep GPUs saturated, accelerating training throughput, efficiency.
Cloud Computing
Unifies data across sites and clouds, enforcing policies and minimizing egress while accelerating cloud workloads.
Data Analytics
Creates one global namespace and orchestrates datasets to compute, speeding queries, pipelines, and interactive analytics.
Machine Learning
Unifies training and inference data, automates placement, and maximizes GPU utilization for faster model iteration.
GPU Acceleration
Feeds GPUs from Tier 0 shared NVMe, eliminating bottlenecks and boosting token throughput and TTFT.
High-performance Computing (HPC)
Delivers parallel file performance across sites, keeping compute saturated while simplifying data movement and access.
Retrieval Technologies (RAG)
Unifies files and objects, accelerating retrieval augmented generation while ensuring governance.
AI/HPC Performance Demonstrated on Azure
Based on recent performance benchmarking completed on a 128-node GPU cluster built on ND H100 v5 instances, Hammerspace demonstrated:
- Aggregate Bandwidth of 1.2 TB/sec
- Linearly performance scalability to 128 nodes
- >99.9% GPU Utilization demonstrated with Megatron LLM Training
AI/HPC Performance, Half the Cost of Managed Lustre
In today’s hybrid, multi-cloud world, AI performance must scale without complexity, unnecessary overprovisioning, or runaway costs. Legacy file systems designed for HPC in the past are optimized to run in the datacenter, not in the cloud.
Hammerspace can scale performance and capacity independently to avoid over-provisioning, runs on standard cloud infrastructure, and operates at less than half the cost of Managed Lustre.
Read the AnalysisHammerspace can scale performance and capacity independently to avoid over-provisioning, runs on standard cloud infrastructure, and operates at less than half the cost of Managed Lustre.
Get Started with Hammerspace on Azure Cloud
Experience faster performance, lower latency, and reduced costs with Hammerspace on Azure Cloud.
Azure Marketplace