Research Engineer, Inference

Normal Computing Corporation•New York City, NY

About The Position

Normal Computing builds silicon that turns thermal noise from an obstacle into a computational resource. Conventional chips spend most of their energy forcing determinism onto physics; ours compute with it. Stochastic, in-memory, asynchronous: the result is 10-100× more AI inference per dollar, per watt. We co-design the full stack: AI-native EDA systems in production with the world's largest semiconductor companies, and the advanced ASICs they make possible. Backed by $85M+ from the world's leading deep-tech investors and built by scientists, engineers, and operators from the labs that built modern computing. Normal works as one team across New York, Silicon Valley, London, Copenhagen, and Seoul. We hire people who want the hardest version of their craft, across every discipline, at every seniority.

Requirements

Deep understanding of large model inference: attention mechanisms, KV cache, long-context decoding, memory bandwidth constraints
Experience with inference optimization: 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, not just in high-level frameworks
Comfort reasoning from first principles about what a novel substrate can do efficiently
Track record of taking ideas from theory to working implementation on real hardware
Strong programming skills in Python and at least one systems language
Collaborative instinct and ability to work across hardware, architecture, and software teams

Nice To Haves

PhD in machine learning, applied mathematics, physics, electrical engineering, or a related field
Exposure to analog or mixed-signal systems, in-memory compute, or non-von-Neumann architectures
Experience working on hardware that did not yet exist when you joined
Publications or open-source work in efficient inference, stochastic algorithms, or novel computing

Responsibilities

Develop algorithms for transformer inference workloads running on stochastic analog processing-with-memory hardware.
Work directly with hardware and architecture teams to shape what the chip can and should compute natively.
Design numerical methods that exploit thermal noise and analog dynamics rather than working around them.
Build evaluation frameworks and benchmarks that characterize algorithm behavior on real hardware or simulation.
Translate insights about model workloads into constraints and opportunities for hardware design.
Prototype and iterate rapidly as hardware evolves from simulation to silicon.