The overarching goal of the proposal is to take a holistic view to make memory disaggregation practical by addressing challenges in the end host, inside the network, and throughout the entire cluster.
Thanks NSF for supporting the proposal and Deep Medhi for championing it. I'd like to also thank … Continue Reading ››
With the advancement of AI in recent years, GPUs have emerged as a popular choice for training deep learning (DL) models on large datasets. To deal with ever-growing datasets, it is also common to run distributed deep learning over multiple GPUs in parallel. Achieving cost-effectiveness and high performance in these clusters relies on … Continue Reading ››
Just over two weeks ago, we wrapped up the Second Asia-Pacific Workshop on Networking (APNet’2018) in the campus of Tsinghua University in Beijing, China. This is my first time co-chairing a workshop (or a conference), and the scale of APNet'2018 -- over 170 registrants -- gave me a new appreciation for the challenges in pulling … Continue Reading ››
Distributed data-parallel queries run on large collections of resources -- within a single datacenter in case of traditional big data queries and across multiple datacenters in case of geo-distributed analytics. The total amount of available resources to a running query changes over time for a variety of reasons, including changes in the fair share of … Continue Reading ››
Today's cellular Radio Access Networks (RAN) are vast, and operators collect a variety of data to better understand and manage these networks. A fundamental challenge in any geo-distributed system -- including cellular RANs -- is the tradeoff between data collection latency and the accuracy of insights derived from the collected data. While the general problem … Continue Reading ››
All credit goes to Anand. Congratulations!!!
Couple more workshop papers got accepted in last few days!
The APNet one is about coflows on optical networks, following my collaborations with Hong and Kai at HKUST (SIGCOMM'16 and SIGCOMM'17), while … Continue Reading ››
Over the course of last few days, we heard back about the acceptance of three workshops papers:
The HotCloud ones deal with networking for and graph … Continue Reading ››
High-throughput, low-latency lock managers are useful for building a variety of distributed applications. A key tradeoff in this context can be expressed in terms of the amount of knowledge available to the lock manager. On the one hand, a decentralized lock manager can increase throughput by parallelization, but it can starve certain categories of applications. … Continue Reading ››
APNet'18 took place in August 2-3, 2018. Read about how it went!
Kun Tan and I are co-chairing for the Second Asia-Pacific Workshop on Networking (APNet’2018) to be held in Beijing, China. APNET aims to bring together the very best researchers in computer networking and systems to a live forum in the … Continue Reading ››