Skip to content

Scientific IT and Application Support#

SCITAS (Scientific IT and Application Support) is the central hub for scientific computing at EPFL, offering advanced computational resources and High-Performance Computing (HPC) expertise. In addition to HPC services, SCITAS provides cutting-edge artificial intelligence (AI) solutions.

Core Services#

High-Performance Computing (HPC) on premise#

  • Supercomputers: HPC/AI clusters equipped with CPU or CPU/GPU nodes.
  • Software Stack: A fine-tuned scientific software stack.
  • Advanced Networking: High-bandwidth, low-latency network infrastructure.
  • Scalable Storage: Petabyte-scale parallel distributed and long-term storage shared across clusters.
  • HPC Software Development: Expert support in software development and optimization to maximize user-code efficiency.

HPC on the Public Cloud#

  • On-Demand Clusters: Users can request exclusive HPC/AI clusters on public cloud providers, such as GCP or Exoscale, tailored to their specific needs.
  • Rapid Deployment: Fully operational clusters are provisioned within less than one day.
  • Flexibility and Scalability: Public cloud solutions offer unmatched adaptability to accommodate unexpected computational demands.

HPC and Artificial Intelligence (AI) Support#

  • AI Computing Resources: Specialized hardware and infrastructure optimized for machine learning, deep learning, and other AI-driven applications.
  • HPC Software Development: Expert support in software development and optimization to maximize user-code efficiency.
  • AI Competency: Collaboration with AI specialists to support algorithm development, training, and model deployment.

Community Services#

  • Customized Training Programs: Hands-on workshops and training sessions in HPC and AI technologies for researchers and students.
  • Expert Advisors: Personalized guidance on project planning, resource allocation, and optimization strategies.
  • Competitive Pricing: Flexible and affordable pricing pay-per-use model.
  • Sustainable Practices: Commitment to green IT through energy-efficient computing and reduced environmental impact.