Infrastructure

• Time on a shared HPC cluster ready for HPC and AI workloads

• Access to a CPUs and GPUs suitable for workload

• HPC workloads include

  • Computational Fluid Dynamics, Computational Solid Mechanics

  • Climate and weather modeling

  • Computational chemistry

  • Biosciences

  • Earth sciences

  • Geosciences

• AI workloads include

  • Pre-training and Transfer learning/fine-tuning of models

    • Natural language processing

    • Computer Vision

    • Speech recognition

    • Machine learning models (decision trees, SVMs,K-means, KNNs etc)

  • Inference

    • Batch inference

• High performance shared parallel filesystem

• Workload manager

• User based accounting

User support

• Ticket based support with a response time based on the SLA

• Optimized software installation

  • Use either pre-installed software by RTS

  • Or RTS helps you install your software

• Run workloads as

  • Scheduled batch jobs

  • Interactive sessions

    • Jupyter Lab

    • VS Code like environment

• Access to

  • selection of common datasets of AI model development

  • examples of common workflows and case files

Training

• On-site and Online workshops providing

  • Technical content from experts in the field

  • Hands-on exercises

  • Time on RTS cloud for practice and prototyping

  • Topics include:

    • HPC essentials and programming (Linux, Git,Python, Containers, MPI, OpenMP, CUDA, Performance & Debugging tools)

    • Modeling and Simulation in various scientific domains

      • Computational Fluid Dynamics, Computational Solid Mechanics

      • Climate and weather modeling

      • Computational chemistry

      • Biosciences

      • Earth sciences

      • Geosciences

    • Machine Learning and Deep Learning topics

      • PyTorch, Tensorflow,

      • DeepSpeed for distributed training (data and model parallelism)

      • Distributed hyperparameter Optimization

      • MLOps

Consulting

• Free 30 min consultation to gather requirements

• Optimized software installation on RTS cloud

• Consulting on how to best use the resources for your workloads

• On-boarding training on how to use RTS cloud

• Code refactoring to accelerate workloads on multiple CPUs and GPUs

Mentoring

• Student internships to gain experience by working on short projects

  • Earn a certificate of completion upon successful conclusion of the project

Collaboration

• Supervision in academic research and co-authorship of publication in international conferences and journals

• Certification programs in collaboration with host Universities

• Industrial collaborative project