Jasmine Mou and Krz Sywula give a moonshot talk on using Machine Learning Optimization Algorithms in HTTP Latency Tuning on Nginx. Jasmine and Krz discuss techniques for dealing with the massive number of tunables to identify the best parameter combinations for different workloads, and streamline the process to minimize human intervention. They focus on optimizing HTTP latency for an Nginx server. In this talk they discuss their methodology and share their experimental results. More info: https://netdevconf.info/0x18/27
Jesse Brandeburg and Kamel Ayari present work on enhancing netdev reviews using AI. I have to admit my bias: This is my favorite submission! You dont need to read between the lines and guess the reviewers bias, it's a machine giving you feedback! Does this mean we are going to get rid of all those pesky reviewers? Come and find out ;-> More info: https://netdevconf.info/0x18/26
Jasmine Mou will give a talk on Machine Learning Practices in Network Traffic across Data Centers. With massive growth of network infrastructure it is getting overwhelming for network engineers in network traffic management and capacity planning. Jasmine will share ByteDance practices of how machine learning, statistical profiling, and visualization techniques are applied with network traffic data to address these challenges. More info: https://netdevconf.info/0x18/16
Anjali Singhai, Shaopeng He and Sridhar Samudrala will give a talk on Advancing TCP with Device Memory[1] and Collective Communication(CC). Anjali, Shaopeng and Sridhar build up on devmem work on the Intel IPU NICs to enable efficient CC across clusters. The authors will present detailed performance data to demonstrate the enhanced TCP’s effectiveness, showcasing its comparability to RDMA and its superiority over standard TCP, particularly in handling larger packet sizes. More info: https://netdevconf.info/0x18/33
cheers, jamal [1] devmen bof https://netdevconf.info/0x18/36