Applications are invited for an internship position at the IMDEA Software Institute in Madrid, Spain.
The successful candidates will join the research lab under the supervision of Dr. Thaleia Dimitra Doudali.
The internship project seeks to identify the most efficient computational infrastructures, reduce resource consumption, and improve scalability while maintaining accuracy and speed in LLM-based applications. It aligns with the broader field of computer systems, artificial intelligence, and high-performance computing (HPC) by addressing the pressing challenge of efficiently deploying large AI models at scale. The project seeks to address the lack of efficient optimization strategies for deploying these models without compromising on performance or accuracy. It involves setting up experiments to benchmark system performance, reviewing current LLM inference methods, and evaluating metrics like latency, throughput, and resource usage. The project will also include testing optimization techniques, such as quantization and pruning, and propose new system architectures for large-scale LLM deployment.
The position requires good programming and problem solving skills and proficiency in spoken and written English. A background on machine learning, LLMs and computer architecture and systems (course work, coding experience etc.) is highly desirable. Programming skills in Python, with familiarity in CUDA and NVIDIA Nsight being valuable. This is a great opportunity for graduating students who consider future PhD studies and want to get hands-on research experience. The project can be shaped in the scope of a thesis (e.g., Masters).
The IMDEA Software Institute is ranked among the best European research institutes in computer science. The institute provides an internationally competitive stipend and support for research related travel. The working language at the institute is English. Knowledge of Spanish is not required.
The position has a duration of 6 months. The starting date is October 1st, 2024. Deadline for applications is September 19th, 2024. Review of applications will begin immediately, and continue until the position is filled.
Applicants interested in the position should submit their application at https://careers.software.imdea.org/ using reference code 2024-09-intern-sysllm.
The recruitment process will comply with the IMDEA Software Institute’s OTM-R Policy.
For any questios about the position, please contact Thaleia Dimitra Doudali directly ().
