Resources

Research Computing

Research Computing

Navigating the HPC & Research Computing Landscape

High-performance computing (HPC) and data-intensive workflows require matching your specific computational needs with the right hardware, financial framework, and security infrastructure. Whether you are training AI/ML models, storing long-term research datasets, running virtual machines, or handling sensitive data under regulatory mandates, the compute ecosystem spans multiple tiers—ranging from on-site hardware and regional clusters to public cloud providers and federally subsidized supercomputers.

 

To help you navigate these options, Mines Research Computing (RC@Mines) supports a broad matrix of resources structured around your budget model, whether you rely on operational expenditures (OpX), capital investments (CapX), federally-subsidized HPC through NSF Advanced Cyberinfrastructure Coordination Ecosystem: Services & Support (ACCESS) or National AI Research Resource (NAIRR), or grants.

 

Use the comparison matrix and selection flowchart below to identify the optimal platform for your research.

Table 1: Mines Research Computing Resource Matrix — An overview of available HPC, AI, long-term storage, virtual machine, and controlled data infrastructure categorized by provider tier and financial framework (OpX, CapX, or Subsidized/Free).

Mines HPC Resource Selection Flow Chart — A decision tree guiding researchers, faculty, and students through selecting the appropriate computing or storage platform based on funding model, computational scale, AI/ML requirements, and data compliance standards.

Next Steps & Consultation Support

Selecting the right environment early in your project timeline prevents unnecessary costs and accelerates your computational output. If you are unsure which path aligns best with your research, or if you need assistance with setup, our team is available to help with:

 

  • Resource & Workflow Optimization: Identifying whether public cloud environments, national centers, regional resources or the  local cluster will fit your software stack.
  • Subsidized Grant Allocations: Securing free computational time and storage grants through ACCESS or NAIRR programs.
  • Compliance & Controlled Data: Navigating security frameworks for regulated datasets requiring compliance.
  • Hardware Procurement: Planning capital equipment purchases for long-term multi-year compute nodes or Petabyte-scale storage.

 

Ready to get started? Reach out directly to RC@Mines to schedule a one-on-one consultation, request account allocations, or discuss customized infrastructure for your lab.

Cloud Services

Google Cloud Platform (GCP)

GCP is a suite of cloud computing services designed for building, deploying, and managing applications at scale. It offers a wide range of solutions, including computing, storage, networking, AI/ML, and security, catering to research workloads. GCP features include: 

  • Key Services: Features include Google Compute Engine, Google Cluster Toolkit, and Google BATCH for HPC workload management. 
  • Global Infrastructure: Ensures high availability and low latency through a network of regions and zones worldwide. 
  • Data and AI Capabilities: Offers BigQuery for analytics and Cloud AI/TensorFlow for advanced machine learning. 
  • Robust Security: Encryption and compliance for secure operations. 
  • Financial Commitment: GCP is a paid service, requiring allocated funds. 

Amazon Web Services (AWS)

AWS is a cloud platform offering a range of on-demand services, including computing, storage, networking, and analytics. 

  • Key Services: Includes Amazon EC2, Amazon S3, Amazon Sagemaker, and Amazon RDS. 
  • Global Infrastructure: Provides low latency and high availability through Regions and Availability Zones worldwide. 
  • AI/Machine Learning Tools: Features Amazon SageMaker for model development, Amazon Bedrock for foundational models, and AWS ParallelCluster for high-performance computing. 
  • Financial Commitment: AWS is a paid service, requiring allocated funds. 

HPC & Data Center Services

Mines Research Computing offers a variety of on-premise and off-premise HPC offerings for researchers

Mines x CU Boulder: Blanca HPC Condo Cluster

Mines has partnered with CU Boulder to integrate our HPC workflows into Blanca—CU’s HPC condo cluster.

Why Choose Blanca for Your Next Project?

  • Seamless Integration: Run your existing Mines HPC workflows effortlessly with support from Mines RC staff.

  • Guaranteed Longevity: Purchase your own compute hardware with an incredible 5+ years of support.

Interested in Using Blanca for your research? Explore the Blanca Rates and Policies today and level up your research infrastructure. Contact the Research Computing team today at rc@mines.edu.

Note: Blanca does not support CUI, ITAR, and other compliance data workloads.

Mines x TACC Partnership for Affordable Node-hour rate HPC!

If you are running massive datasets, simulations, or heavy machine learning models, world-class compute power is now within your reach.

Why Bring Your Workloads to TACC?

  • Unprecedented Scale: Effortlessly deploy your heaviest, most demanding CPU and GPU workloads on world-tier architecture.

  • Exclusive Node-Hour Rates: Maximize your research budget with special, highly competitive pricing negotiated for the Mines community.

  • Compliance: TACC is capable of supporting workloads requiring CUI, ITAR, and other strict regulatory compliance rules.

  • Global-Scale Breakthroughs: Stop compromising on model size or resolution. Run your code on the infrastructure that powers global discovery.

From advanced materials modeling to massive geospatial data processing, TACC gives you the runway to innovate without limits.

Wendian

The Wendian computing cluster, provided by Research Computing (RC), offers campus researchers and their collaborators a high-performance computing (HPC) environment. This Linux cluster features high-speed, low-latency networking and a high-speed parallel file system.  

Key features of Wendian include:

  • High-speed, low-latency networking. 
  • High-speed parallel file systems. 
  • Pools of a variety of types of compute nodes: standard, high-memory, and GPU. 
  • Parallel processing across cores, across nodes, and across multiple GPUs. 
  • Dedicated personal and group disk space and extensive temporary scratch space for active use. 
  • Research Computing staff provide sysadmin and consulting support and training for users of the system. 

Wendian is a pay-per-use machine. For details on up-to-date pricing, please see our policies page. 

Compute Partition:

Designed for research projects that have funding for operation use.  

Student Partition:

Designed for classroom use, student projects, and undergraduate research, the student partition offers access to 7 high-performance, Intel-based compute nodes. For details and access eligibility, please review our student partition guidelines on our documentation site. 

Mio

Mio is the oldest operational high-performance computing (HPC) system at the Colorado School of Mines, boasting a 53+ Tflop capacity. Mio no longer accepts new purchases on the system. 

  • Students can access Mio through previous advisor-purchased nodes, or for class projects, with Research Computing staff support available via consultations and workshops. 

For more details, please see our Mio documentation. 

Data Center Services

The CTLM Data center provides research infrastructure hosting, including private partitions on Wendian and non-managed servers in the hosted area. For more information, please see our Data Center guidelines. 

Orebits

OreBits is an on-premise storage platform designed to meet the growing data storage needs of the Mines community. OreBits provides a balanced solution, offering excellent performance, robust data protection, and a competitive price point. 

Ready to learn more? Check out our Orebits pricing and documentation pages for more details.  

Interested in expanding your group’s storage capacity? Submit a Research Computing Service Request. 

Data Management Services

Globus

Globus is a service designed to provide high-performance, secure, and reliable data transfer and management.  Globus is a convenient way to transfer files between users and other sites. It has both a web interface and a command line interface (CLI). For more information, please take a look at our Globus documentation page. 

Research Data Management (Mines Library)

Research data management (RDM) involves making research data accessible and reusable, often required by funders.  

  • Data repositories are key for disseminating research data. Access control mechanisms like passwords, encryption, and permissions systems are used to manage data sharing. 
  • Mines offers an institutional repository for end-of-project data deposits, helping researchers meet public accessibility requirements. It supports various access levels, multiple file types, and small to medium-sized datasets. 
  • Researchers can also use external repositories (subject-specific, national, international) and those offered by journals/societies. 
  • External repositories have varying requirements for data domain, reuse, access, file formats, and metadata. Researchers must ensure chosen repositories comply with funder guidelines. 

A data management plan (DMP) is crucial; see https://libguides.mines.edu/RDM for more information.