Scalable Cube

Scalable Cube is an enterprise ready, supported distribution of Cubeman (ClUster Batch Environment MANagement), an open-source, workload scheduler that supports a wide variety of HPC and analytic applications. Whether deployed on-site, on virtual infrastructure, or in the cloud, customers can take advantage of top-quality support services from HPC Scalable, helping ensure the success of managing their HPC workloads.

HPC Scalable provides fully tested binaries optimized for ease of installation, commercial support, training, and consulting services for Scalable Cube.


 
Download open source Cube
or Request Scalable Cube Evaluation


Scalable Cube offers advanced features including fairshare, preemption, job arrays, MPI integrations, job dependencies etc. It is production proven in environments with thousands of compute nodes, hundred thousands of cores and tens of millions of jobs.

Download the Scalable Cube datasheet to learn more about its advanced capabilities.


 

Scalable Cube 2.0 New Features

FeaturesDescriptionBenefits
Enterprise scheduling capabilities • Fairshare
• Preemption
• Ownership
• Flexible resource limits
Manage workload in large mission critical environment
Robust, scalable, and high performance• Fully automated failover
• Scale to thousands of hosts
• Can schedule millions of short running jobs
Suitable for traditional HPC workloads, high throughput workloads, and modern data analytics workloads
Built-in GPU scheduling• Automatically detect available GPUs
• Schedule jobs based on GPU resources
• Support GPU affinity
Maximize GPU resource utilization and workload throughput
Docker support• Running jobs inside Docker containersInsulate job environment for security and performance
NUMA support• Support job CPU affinityEnsure job performance in NUMA environment
ssh X11 forward support• Support ssh X11 forward across different Linux distributions with VNCEnable remote desktop deployment for ease of management and data security
Enhanced Web GUI• Web GUI supporting Scalable Cube and other workload managementLower learning curve for users when using the compute cluster

Support Programs

 Standard SupportPremium Support
Hours of coverageLocal business hours 9am - 5pmLocal business hours 9am - 5pm
with 24x7 for Severity 1 issues
Communication methodEmailEmail and Phone
Number of casesEmail and PhoneEmail and Phone
Response TimeInitial/OngongInitial/Ongong
Severity 12/2 business hours1/1 hour
Severity 24/4 business hours2/4 business hours
Severity 31/1 business day4 business hours/1 business day
Roadmap request priorityNoYes
Developer accessNoYes