Current Articles
- AMD Helios vs. NVIDIA Vera Rubin NVL72A side-by-side comparison of two 72-GPU rack architectures.
- AMD Helios rack-scale systemInstinct GPUs, EPYC CPUs, Pensando networking, ROCm, and the questions buyers should test.
- NVIDIA Vera Rubin NVL72The integrated rack architecture that sets the current comparison baseline.
- Etched inference cluster reviewWhat Etched means by an inference cluster, where Linux fits, and why its performance case still needs independent proof.
Linux Compute Clusters Book
- Chapter 1: Introduction to Clusters
- Appendix 1: Inference Clusters
- Appendix 2: Distributed AI Training Clusters
- Chapter 2: Cluster Node Design
- Chapter 3: Cluster Interconnects
- Chapter 4: Storage for a Linux Compute Cluster
- Chapter 5: Power, Cooling and Physical Design
- Chapter 6: Linux Software and Parallel Programming Environment
- Chapter 7: Scheduling, Resource Management, and Performance
- Chapter 8: Cluster Administration, Automation, and Security
- Chapter 9: Running a Production Cluster Service
- Chapter 10: Building a Cluster from the Ground Up
- Chapter 11: Commercial Application: Seismic Imaging