This talk describes the design of shadow cache, a lightweight component to track the working set size of Alluxio cache. Shadow cache can keep track of the working set size over the past window dynamically, and is implemented by a series of bloom filters. We’ve deployed the shadow cache in Facebook Presto and leverage the result to understand the system bottleneck and help with routing design decisions.
ALLUXIO DAY VI 2021
October 12, 2021
This talk describes the design of shadow cache, a lightweight component to track the working set size of Alluxio cache. Shadow cache can keep track of the working set size over the past window dynamically, and is implemented by a series of bloom filters. We’ve deployed the shadow cache in Facebook Presto and leverage the result to understand the system bottleneck and help with routing design decisions.
Video:
Presentation Slides:
Complete the form below to access the full overview:
Videos
TorchTitan is a proof-of-concept for Large-scale LLM training using native PyTorch. It is a repo that showcases PyTorch's latest distributed training features in a clean, minimal codebase.
In this talk, Tianyu will share TorchTitan’s design and optimizations for the Llama 3.1 family of LLMs, spanning 8 billion to 405 billion parameters, and showcase its performance, composability, and scalability.
As large-scale machine learning becomes increasingly GPU-centric, modern high-performance hardware like NVMe storage and RDMA networks (InfiniBand or specialized NICs) are becoming more widespread. To fully leverage these resources, it’s crucial to build a balanced architecture that avoids GPU underutilization. In this talk, we will explore various strategies to address this challenge by effectively utilizing these advanced hardware components. Specifically, we will present experimental results from building a Kubernetes-native distributed caching layer, utilizing NVMe storage and high-speed RDMA networks to optimize data access for PyTorch training.