Average Ratings 0 Ratings
Average Ratings 0 Ratings
Description
Hammerspace innovatively leverages the local NVMe storage embedded within GPU servers, converting it into a high-performance, shared storage tier designed specifically for large-scale AI training and checkpointing workloads. This approach eliminates bottlenecks inherent in legacy storage systems that struggle to keep GPUs fully utilized, while significantly reducing power consumption and external storage expenses. The platform’s parallel file system architecture supports massive scalability, allowing data to be served simultaneously to thousands of GPU nodes with minimal latency. Hammerspace integrates seamlessly with existing Linux storage servers and supports hybrid cloud environments, enabling data orchestration between on-premises and cloud infrastructure. It delivers record-setting performance validated by MLPerf benchmarks, proving its efficiency for demanding machine learning workloads. Customers such as Meta and Los Alamos National Laboratory trust Hammerspace to optimize their AI data pipelines and infrastructure investments. With quick setup and intuitive management, Hammerspace helps organizations accelerate AI projects while reducing operational complexity. By transforming underutilized storage into a powerful resource, Hammerspace drives cost savings and faster innovation.
Description
Create a robust NVMe over Fabrics high-performance shared storage solution with MayaScale that allows for the integration of directly attached NVMe resources into a unified storage pool. This solution enables the flexible provisioning of NVMe namespaces to clients who require high performance with minimal latency. After usage, clients have the option to return NVMe storage back to the shared pool, eliminating issues associated with over-provisioning or unutilized NVMe storage typical of direct-attached setups. The network-agnostic architecture employs RDMA for on-premises deployments and standard TCP for cloud environments, ensuring versatility. Clients can access true NVMe devices using a conventional NVMe driver stack, negating the need for any proprietary drivers. You can easily configure and implement NVMe over Fabrics SAN infrastructure at rack scale in your data center by aggregating diverse NVMe devices through RDMA-compatible connections, such as ROCE, iWARP, or Infiniband. Furthermore, even in public cloud settings, users can harness the benefits of NVMe over Fabrics via the standard TCP/IP protocol, which eliminates the requirement for specialized RDMA hardware or SRIOV virtualization. This innovative approach optimizes resource utilization while maintaining high performance across various deployment scenarios.
API Access
Has API
API Access
Has API
Integrations
Google Cloud Platform
Microsoft Azure
AWS Marketplace
Amazon
Amazon RDS
Amazon Web Services (AWS)
Azure Marketplace
MongoDB
MySQL Workbench
Oracle Cloud Infrastructure FastConnect
Integrations
Google Cloud Platform
Microsoft Azure
AWS Marketplace
Amazon
Amazon RDS
Amazon Web Services (AWS)
Azure Marketplace
MongoDB
MySQL Workbench
Oracle Cloud Infrastructure FastConnect
Pricing Details
No price information available.
Free Trial
Free Version
Pricing Details
No price information available.
Free Trial
Free Version
Deployment
Web-Based
On-Premises
iPhone App
iPad App
Android App
Windows
Mac
Linux
Chromebook
Deployment
Web-Based
On-Premises
iPhone App
iPad App
Android App
Windows
Mac
Linux
Chromebook
Customer Support
Business Hours
Live Rep (24/7)
Online Support
Customer Support
Business Hours
Live Rep (24/7)
Online Support
Types of Training
Training Docs
Webinars
Live Training (Online)
In Person
Types of Training
Training Docs
Webinars
Live Training (Online)
In Person
Vendor Details
Company Name
Hammerspace
Founded
2018
Country
United States
Website
hammerspace.com
Vendor Details
Company Name
ZettaLane Systems
Founded
2018
Website
www.zettalane.com/maya-nvmeof-linux-rdma-tcp.html
Product Features
Data Management
Customer Data
Data Analysis
Data Capture
Data Integration
Data Migration
Data Quality Control
Data Security
Information Governance
Master Data Management
Match & Merge
Disaster Recovery
Administration Policies
Bare-Metal Recovery
Encryption
Failover Testing
Flexible Data Capture
Multi-Platform Support
Multiple Data Type Support
Offline Storage