Skip to main content

Unlock Marlowe’s Potential for Breakthrough Research

Marlowe, one of Stanford’s high-performance computing (HPC) clusters, delivers GPU-powered computing, equipping researchers with the resources required for complex data science and artificial intelligence research.

To help faculty members new to Marlowe accelerate their research, we are offering 5,000 free GPU hours to get started. Researchers who apply to use Marlowe will also receive a discounted rate. 

Marlowe Enables Data-Driven Discovery

Marlowe is an NVIDIA DGX H100 Superpod, constructed with NVIDIA’s reference architecture and designed for cutting-edge computational performance. It comprises 31 NVIDIA H100 nodes, collectively providing 248 NVIDIA H100 GPUs and 2.5PB of high-performance DDN Lustre storage.

Each NVIDIA DGX H100 node includes:

  • GPU: 8x NVIDIA H100 80GB GPUs
  • CPU: 2x Intel Xeon Platinum 8480C CPUs (112 cores/node)
  • Memory: 2TB of RAM
  • NVSwitch: 4x NVLink connections, providing up to 900 GB/s GPU-to-GPU bandwidth
  • Node-local Storage: 30TB NVMe
  • Networking: 8x 400Gbps NDR InfiniBand connections, providing up to 3.2Tbps bandwidth 

Marlowe’s powerful computing system will accelerate complex simulations, AI training, and interdisciplinary breakthroughs, advancing Stanford’s global research impact.

Ready to leverage Marlowe for your research? Submit an application for access. 

Support

The Marlowe Research Data Science team is available to aid faculty, graduate students, researchers, and postdoctoral students in maximizing Marlowe. 

The team supports various functions, including: 

  • Designing tools to achieve optimal system performance.
  • Collaborating with research groups and students to educate them on techniques and methods for optimizing usage.
  • Providing software tools and operating workflows that enable groups to follow best open science practices seamlessly.

Additional resources

Share Feedback

DISCLAIMER: UIT News is accurate on the publication date. We do not update information in past news items. We do make every effort to keep our service information pages up-to-date. Please search our service pages at uit.stanford.edu/search.