Research Intern - LLM Inference Acceleration and Optimization
 Microsoft | |
 United States, Washington, Redmond | |
 Nov 23, 2024 | |
|
OverviewResearch Internships at Microsoft provide a dynamic environment for research careers with a network of world-class research labs led by globally-recognized scientists and engineers, who pursue innovation in a range of scientific and technical disciplines to help solve complex challenges in diverse fields, including computing, healthcare, economics, and the environment.If you are excited about investigating and implementing cutting-edge large language model (LLM) inference techniques and optimizations like quantized KV-caches, flash/paged/radix attention, speculative decoding, and advanced collective communication on graphics processing units (GPUs), come join the AIFX team at Microsoft Azure and contribute to a production-focused, planetary-scale LLM serving stack that is being built on top of excellent open-source efforts like vLLM, SGLang, and HuggingFace. The work includes investigation of cutting-edge, state-of-the-art approaches like "You only cache once (YOCO)" and leveraging them to save memory and compute for serving LLMs at scale. You will get a chance to explore, implement, optimize, and publish your research ideas in collaboration with teams at Microsoft working on real-world production workloads at an unprecedented scale.
ResponsibilitiesResearch Interns put inquiry and theory into practice. Alongside fellow doctoral candidates and some of the world's best researchers, Research Interns learn, collaborate, and network for life. Research Interns not only advance their own careers, but they also contribute to exciting research and development strides. During the 12-week internship, Research Interns are paired with mentors and expected to collaborate with other Research Interns and researchers, present findings, and contribute to the vibrant life of the community. Research internships are available in all areas of research, and are offered year-round, though they typically begin in the summer. | |