CUDA Graphs

Includes some useful examples below. 

Enabling Dynamic Control Flow in CUDA Graphs with Device Graph Launch

By Sally Stevenson, Stephen Jones and Fred Oh

CUDA Graphs significantly reduce the overhead of launching a large batch of user operations by defining them as a task graph, which may be launched in a single operation. Knowing the workflow upfront enables the CUDA driver to apply various optimizations, which cannot be performed when launching through a stream model. 

However, this performance comes at the cost of flexibility: if the full workflow is not known in advance, then GPU execution must be interrupted to return to the CPU to make a decision.

CUDA device graph launch solves this problem by enabling a task graph to be performantly launched from a running GPU kernel, based on data that is determined at run time. CUDA device graph launch offers two distinct launch modes—fire and forget, and tail launch—to enable a wide range of applications and use.

This post demonstrates how to use device graph launch and the two launch modes. It features the example of a device-side work scheduler, which decompresses files for data processing.  ... ' 

Robustly Labeled Graphs

Interesting and Technical piece in the Google AI Blog. 

Robust Graph Neural Networks  in The Google Blog

Tuesday, March 8, 2022

Posted by Bryan Perozzi, Research Scientist and Qi Zhu, Research Intern, Google Research

Graph Neural Networks (GNNs) are powerful tools for leveraging graph-structured data in machine learning. Graphs are flexible data structures that can model many different kinds of relationships and have been used in diverse applications like traffic prediction, rumor and fake news detection, modeling disease spread, and understanding why molecules smell.

As is standard in machine learning (ML), GNNs assume that training samples are selected uniformly at random (i.e., are an independent and identically distributed or “IID” sample). This is easy to do with standard academic datasets, which are specifically created for research analysis and therefore have every node already labeled. However, in many real world scenarios, data comes without labels, and labeling data can be an onerous process involving skilled human raters, which makes it difficult to label all nodes. In addition, biased training data is a common issue because the act of selecting nodes for labeling is usually not IID. For example, sometimes fixed heuristics are used to select a subset of data (which shares some characteristics) for labeling, and other times, human analysts individually choose data items for labeling using complex domain knowledge.  .... '