Gang Scheduling Performance Benefits for Fine-Grain Synchronization

By Dror G. Feitelson, Larry Rudolph. 1992

Subfields: ML Systems & Hardware

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In ML Systems & Hardware

1. Sanjay Ghemawat
2. Michael Isard
3. Aakanksha Chowdhery
4. S. Hand
5. Joseph Dean
6. Hyeontaek Lim
7. Ion Stoica
8. Amir Salman Avestimehr

Citations in the corpus

• Pathways: Asynchronous Distributed Dataflow for ML (2022, background): “Even for GPUs or other accelerators that can execute concurrent computations, gang scheduling allows more efficient execution of collectives (Feitelson and Rudolph, 1992).”
• Quincy: fair scheduling for distributed computing clusters (2009, methodology): “This is the primary reason that new approaches such as the scheduling framework described in this paper are needed, and standard techniques such as gang scheduling [ 10 ] are not applicable to our setting.”
• Parallel Job Scheduling - A Status Report (2004, background): “Gang scheduling was originally developed in order to support fine-grain synchronization of parallel applications [19].”