Postdoctoral Research Associate - Programming Systems for Extreme Heterogeneity

Overview

The Computer Science and Mathematics Division at Oak Ridge National Laboratory (ORNL) is seeking an exceptional Postdoctoral Researcher to advance programming systems for extreme-scale, heterogeneous computing. This role sits at the intersection of compiler infrastructure, formal verification, runtime systems, and hardware/software co-design—expanding into emerging paradigms such as quantum and analog computational models.

You will explore how compilers, runtimes, and AI-driven agents can co-optimize complex architectures, reasoning across conventional processors (CPUs/GPUs), custom accelerators, and next-generation quantum and analog devices. Your work will contribute to a vision of autonomous system engineering in which AI agents understand algorithmic intent, architectural constraints, and performance models—driving system co-design across the entire computing stack.

Major Duties/Responsibilities

• LLVM Compiler Innovation: Advance the LLVM toolchain to support autonomous "closed-loop" optimization environments, enabling AI agents to analyze, transform, and validate IR with performance-driven reasoning.
• HPC System Co-Design: Investigate methods through which AI agents guide architectural design decisions for massively parallel heterogeneous systems, including pathways that integrate quantum or analog co-processors.
• Performance Engineering at Scale: Profile, model, and optimize distributed codes using MPI, OpenMP, CUDA, ROCm, and related HPC technologies while bridging theoretical AI models with real hardware constraints.
• Cross-Paradigm Integration (new optional emphasis): Explore how compilers and runtimes can unify classical, quantum, and analog execution models under a shared optimization framework.

Basic Qualifications

• Ph.D. in Computer Science, Computer Engineering, or a closely related field.
• Compiler and Systems Expertise: Strong experience with LLVM (e.g., writing passes, IR transformations) and HPC programming in C++, MPI, OpenMP, CUDA, or related models.
• Solid understanding of distributed-memory systems and many-core architectures.

Preferred Qualifications

• Formal Methods: Experience with automated reasoning or verification tools (e.g., Z3, TLA+, LEAN) to ensure correctness of compiler- or AI-generated code.
• Mathematical Foundations: Background in linear algebra, graph algorithms, and related areas central to dataflow and tensor optimization.
• Hardware/Software Co-Design: Familiarity with architectural simulators (e.g., Gem5, SST) or HDLs.
• Quantum / Analog Computing: Exposure to quantum programming models (e.g., QIR, Q#, Qiskit, Cirq) or analog/neuromorphic systems; interest in integrating these paradigms into unified compilation and optimization strategies.
• Programming Languages & Frameworks: Experience with Julia, Rust, Python, Kokkos, OpenSHMEM, or similar emerging ecosystem tools.

Special Requirements

Applicants cannot have received their Ph.D. more than five years prior to the date of application and must complete all degree requirements before starting their appointment. The appointment length will be up to 24 months with the potential for extension. Initial appointments and extensions are subject to performance and availability of funding.