Average Ratings 0 Ratings

Total
ease
features
design
support

No User Reviews. Be the first to provide a review:

Write a Review

Average Ratings 0 Ratings

Total
ease
features
design
support

No User Reviews. Be the first to provide a review:

Write a Review

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 No 

API Access

Has API No 

Screenshots View All

Screenshots View All

Integrations

Google Cloud Platform Yes 
Microsoft Azure Yes 
AWS Marketplace No 
Amazon No 
Amazon RDS No 
Amazon Web Services (AWS) Yes 
Azure Marketplace Yes 
MongoDB No 
MySQL Workbench No 
Oracle Cloud Infrastructure FastConnect No 
Red Hat CloudForms No 

Integrations

Google Cloud Platform Yes 
Microsoft Azure Yes 
AWS Marketplace Yes 
Amazon Yes 
Amazon RDS Yes 
Amazon Web Services (AWS) No 
Azure Marketplace No 
MongoDB Yes 
MySQL Workbench Yes 
Oracle Cloud Infrastructure FastConnect Yes 
Red Hat CloudForms Yes 

Pricing Details

No price information available.
Free Trial No 
Free Version No 

Pricing Details

No price information available.
Free Trial No 
Free Version No 

Deployment

Web-Based Yes 
On-Premises No 
iPhone App No 
iPad App No 
Android App No 
Windows No 
Mac No 
Linux No 
Chromebook No 

Deployment

Web-Based Yes 
On-Premises Yes 
iPhone App No 
iPad App No 
Android App No 
Windows No 
Mac No 
Linux No 
Chromebook No 

Customer Support

Business Hours No 
Live Rep (24/7) No 
Online Support Yes 

Customer Support

Business Hours Yes 
Live Rep (24/7) No 
Online Support Yes 

Types of Training

Training Docs Yes 
Webinars No 
Live Training (Online) Yes 
In Person No 

Types of Training

Training Docs Yes 
Webinars Yes 
Live Training (Online) No 
In Person Yes 

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 No 
Data Analysis No 
Data Capture No 
Data Integration No 
Data Migration No 
Data Quality Control No 
Data Security No 
Information Governance No 
Master Data Management No 
Match & Merge No 

Disaster Recovery

Administration Policies No 
Bare-Metal Recovery No 
Encryption No 
Failover Testing No 
Flexible Data Capture No 
Multi-Platform Support No 
Multiple Data Type Support No 
Offline Storage No 

Product Features

Alternatives

Iotamine Reviews

Iotamine

Iotamine Cloud Private Limited

Alternatives

xiRAID Reviews

xiRAID

Xinnor
Actifio Reviews

Actifio

Google
NVMesh Reviews

NVMesh

Excelero
MayaScale Reviews

MayaScale

ZettaLane Systems
NVIDIA DOCA Reviews

NVIDIA DOCA

NVIDIA