Research Engineer, Algorithms

Job Title

Research Engineer, Algorithms

Role Summary

Develop and validate algorithms for large-language-model (LLM) inference on in-memory compute hardware. Work across hardware, architecture, and software teams to co-design numerical methods, benchmarks, and prototypes from simulation to FPGA and silicon.

Experience Level

Mid-level. No specific years-of-experience specified.

Responsibilities

Primary responsibilities focus on algorithm design, hardware co-design, and evaluation for LLM inference on novel in-memory substrates.

• Develop algorithms for LLM inference that exploit in-memory compute and validate improvements in simulation and on hardware.
• Collaborate with hardware and architecture teams to influence chip capabilities and constraints.
• Design numerical methods that leverage analog dynamics and stochastic behavior.
• Create evaluation frameworks and benchmarks to characterize performance on simulators, FPGAs, and silicon.
• Translate workload characteristics into hardware constraints and opportunities.
• Rapidly prototype and iterate as hardware progresses from simulation to silicon.
• Optimize performance at both algorithmic and low-level implementation layers.

Requirements

Core technical skills and experience expected; additional items listed as nice-to-have.

Must-have

• Deep understanding of LLM inference (attention mechanisms, KV cache, long-context decoding, memory bandwidth constraints).
• Experience with inference optimization such as quantization, sparsity, kernel fusion, or memory-efficient attention.
• Familiarity with stochastic systems, probabilistic methods, numerical analysis, or analog computation.
• Experience implementing algorithms close to hardware (beyond high-level frameworks) and validating on simulators or prototypes.
• Strong programming skills in Python and at least one systems language.
• Ability to work collaboratively across hardware, architecture, and software teams.

Nice-to-have

• Exposure to analog or mixed-signal systems, in-memory compute, or non-von-Neumann architectures.
• Experience working on projects where hardware was under development during algorithm work.
• Publications or open-source work in efficient inference, stochastic algorithms, or novel computing.

Education Requirements

PhD is listed as a bonus/preferred qualification (machine learning, applied mathematics, physics, electrical engineering, or a related field). No specific degree requirement is stated as mandatory.

About the Company

Company: Normal Computing

Headquarters: New York, NY, USA

Normal Computing develops software and hardware solutions for the semiconductor and AI infrastructure industries, specializing in ASICs for image and video diffusion inference and AI accelerators. The company focuses on architecture and microarchitecture of compute blocks, PE array design, ISA co-design, and FPGA prototyping.