Grid Engine
HPCWorks Grid Engine, originally developed by Sun Microsystems and now owned and sold by Siemens, is a distributed resource management (DRM) system designed to optimize workloads and computing resources across high-performance computing (HPC) clusters.
Publisher review
HPCWorks Grid Engine, originally developed by Sun Microsystems and now owned and sold by Siemens, is a distributed resource management (DRM) system designed to optimize workloads and computing resources across high-performance computing (HPC) clusters. It is intended for organizations in fields such as life sciences, manufacturing, and energy that need to run thousands of commercial and open-source applications—including AI and GPU-accelerated workloads—on premises, in the cloud, or in hybrid environments. The software has been proven in clusters with over one million cores or vCPUs, and it supports Linux, Windows, and other OS distributions on x86, Power, and Arm architectures.
The system automatically manages workloads by maximizing shared resources, using policies for priority, utilization, quotas, and limits to align cluster behavior with business objectives. It includes a License Orchestrator to share expensive application licenses and reduce costs. HPCWorks Grid Engine provides comprehensive monitoring and reporting to track resource utilization, and it can deploy and scale dedicated or hybrid HPC clusters in the cloud. It also supports newer frameworks and maximizes GPU resource usage both on-premises and in the cloud.
In the HPC workload manager market, HPCWorks Grid Engine competes directly with Slurm (Simple Linux Utility for Resource Management), which is open source and widely used in academic and government HPC centers, and with Univa Grid Engine, which was originally a fork of the same codebase. Siemens positions HPCWorks Grid Engine as a commercial, enterprise-grade solution with dedicated support and integration with its broader HPCWorks suite (Navops, Insight Pro). Slurm offers no licensing cost but requires more in-house expertise, while Univa Grid Engine is now owned by Altair and rebranded as Altair Grid Engine.
Honest trade-offs include that HPCWorks Grid Engine is a paid product with pricing available only upon request, which can be a barrier for smaller teams or academic groups. Its proprietary nature means users cannot modify the source code, unlike Slurm. The system is complex to configure and tune for specialized workflows, and while it supports many applications, some niche scientific software may require custom integration. Additionally, the documentation and community support are not as extensive as Slurm's open-source ecosystem.
How it works
-
Distributed resource management
Optimizes distributed workloads and computing resources to improve throughput and performance across clusters with over one million cores.
-
Workload automation
Automatically manages workloads by maximizing shared resources, using priority policies, quotas, and limits to meet business objectives.
-
GPU resource support
Supports new frameworks and maximizes use of GPU resources both on-premises and in the cloud for AI and compute-intensive tasks.
-
Cloud cluster deployment
Easily deploys and scales dedicated and hybrid HPC clusters in the user's choice of cloud provider.
-
Comprehensive reporting
Tracks and measures resource utilization with comprehensive monitoring and reporting tools for capacity planning and cost analysis.
-
License Orchestrator
Shares expensive application licenses across the cluster to optimize software asset utilization and reduce costs.
-
Multi-OS and architecture support
Supports Linux, Windows, and other OS distributions on x86, Power, and Arm systems for broad hardware compatibility.
Strengths and trade-offs
Strengths
- Proven in clusters with over one million cores or vCPUs, demonstrating extreme performance at scale.
- Supports thousands of commercial and open-source applications across life sciences, manufacturing, and energy sectors.
- Includes a License Orchestrator that shares expensive application licenses, reducing software costs for organizations.
- Offers comprehensive monitoring and reporting tools that track resource utilization and help optimize capacity planning.
Trade-offs
- Pricing is not publicly listed and requires contacting sales, making it difficult for small teams to budget.
- Proprietary software with no access to source code, limiting customization compared to open-source alternatives like Slurm.
- Configuration and tuning for specialized workflows can be complex, requiring significant expertise and time.
- Community support and documentation are less extensive than Slurm's open-source ecosystem, potentially slowing troubleshooting.
Pricing context
Pricing is not publicly disclosed and is available only upon request by contacting Siemens sales.
Getting started with Grid Engine
-
Request a sales quote
Contact Siemens sales through the HPCWorks Grid Engine website to request pricing and licensing details. Provide your organization's cluster size, workload types, and deployment preferences to receive a tailored quote and trial access.
-
Install the Grid Engine
Download the HPCWorks Grid Engine installer from the Siemens portal after receiving access. Run the installation on your cluster's master node, following the provided documentation for your operating system and architecture (Linux, Windows, x86, Power, or Arm).
-
Configure resource policies
Define workload management policies in the Grid Engine configuration files. Set priority levels, utilization targets, quotas, and limits to align cluster behavior with your business objectives, such as maximizing GPU usage or sharing application licenses.
-
Submit a test job
Use the qsub command to submit a sample job to the Grid Engine scheduler. Specify resource requirements like CPU cores, memory, and GPU count. Monitor the job's execution using qstat to verify that the system correctly allocates resources and applies your policies.
-
Schedule recurring reports
Set up automated monitoring and reporting through the Grid Engine's reporting tools. Configure daily or weekly reports on resource utilization, job completion rates, and license usage to track performance and inform capacity planning decisions.
Frequently Asked Questions
What is Grid Engine and what does it do?
Grid Engine is a distributed resource management system from Siemens that optimizes workloads across HPC clusters. It automates resource sharing, prioritizes jobs, and supports AI and GPU tasks on premises, in the cloud, or hybrid environments.
How does Grid Engine compare to Slurm?
Grid Engine is a commercial product with dedicated support, while Slurm is open source and free. Slurm requires more in-house expertise and has a larger community, but Grid Engine offers enterprise features like License Orchestrator and integration with Siemens HPCWorks suite.
What is Grid Engine pricing and how can I get it?
Grid Engine pricing is not publicly listed and is only available by contacting Siemens sales directly. This can be a barrier for smaller teams or academic groups who need upfront cost information for budgeting purposes.
Does Grid Engine support GPU workloads for AI?
Yes, Grid Engine supports new frameworks and maximizes GPU resource usage both on-premises and in the cloud. It is designed for AI and compute-intensive tasks, making it suitable for life sciences, manufacturing, and energy sectors.
Can Grid Engine deploy clusters in the cloud?
Yes, Grid Engine easily deploys and scales dedicated or hybrid HPC clusters in the user's choice of cloud provider. It supports multi-OS and architecture environments, including Linux, Windows, x86, Power, and Arm systems.
What are the main weaknesses of Grid Engine?
Grid Engine is proprietary with no source code access, limiting customization. Configuration for specialized workflows is complex, and community support is less extensive than Slurm. Pricing requires contacting sales, which can be difficult for small teams.
Alternatives
How Grid Engine compares
Direct head-to-head against 3 competitors. Picked by 7wData.
Grid Engine
- Pricing
- Pricing is not publicly disclosed and is available only upon request by contacting Siemens sales.
- Target
- HPCWorks Grid Engine, originally developed by Sun Microsystems and now owned and sold by Siemens, is a distributed resource management (DRM) system designed to optimize
- Strength
- Proven in clusters with over one million cores or vCPUs, demonstrating extreme performance at scale.
- Watch for
- Pricing is not publicly listed and requires contacting sales, making it difficult for small teams to budget.
Altair Grid Engine
- Pricing
- Contact sales for pricing
- Target
- HPC, life sciences, manufacturing, academic clusters
- Deployment
- On-premise, hybrid
- Strength
- Mature HPC job scheduling with deep resource management
- Watch for
- Complex configuration; Oracle acquisition of Sun Grid Engine legacy
Univa Grid Engine
- Pricing
- Contact sales for pricing
- Target
- Enterprise HPC, cloud bursting, containerized workloads
- Deployment
- On-premise, cloud
- Strength
- Cloud bursting and container support for hybrid HPC
- Watch for
- Acquired by Altair in 2020; integration uncertainty
Apache Mesos
- Pricing
- Free and open source
- Target
- Large-scale distributed systems, data center resource management
- Deployment
- On-premise
- Strength
- Two-level scheduling for diverse workloads (e.g., Hadoop, Spark)
- Watch for
- Steep learning curve; declining community activity since 2020
User reviews
No user reviews yet. Be the first to write one.
Sources
Reporting on this tool draws on these publicly available sources.