Making Serverless Computing Efficient and Scalable
Centre for Networked Intelligence, IISc via YouTube
AI Product Expert Certification - Master Generative AI Skills
Free AI-powered learning to build in-demand skills
Overview
Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Attend this technical seminar exploring advanced approaches to making serverless computing more efficient and scalable. Learn about SPRIGHT, an innovative framework that exploits event-driven shared memory processing to improve serverless data plane efficiency by eliminating data copies and avoiding unnecessary protocol processing overheads. Discover how extended Berkeley Packet Filter (eBPF) enables true event-driven processing to replace heavyweight sidecar proxies typically used in serverless computing. Explore SURE, a unikernel-based serverless framework designed to address secure data plane requirements and fast-function startup through distributed zero-copy communication via seamless interaction between zero-copy protocol stack (Z-stack) and local shared memory processing. Understand how Intel's Memory Protection Keys (MPK) provide lightweight capability for safe shared memory data plane access and library-based sidecar implementation. Examine recent extensions using RDMA to achieve multi-node shared memory communication for even higher performance and lower latency in serverless computing frameworks. The presentation covers fundamental concepts of serverless computing as a fast-growing cloud computing paradigm that simplifies development, deployment, and automated management of modular software functions while exploiting the microservices paradigm for efficient, low-cost computing.
Syllabus
Time: 4:00 PM - 5:00 PM IST
Taught by
Centre for Networked Intelligence, IISc