Research Engineer, In-Memory AI Inference Algorithms

Normal Computing Corporation; Palo Alto (CA); On-site; USD 150,000 - 210,000; Full time

Job summary

Normal Computing is seeking a skilled algorithm and hardware co-design engineer to advance in-memory AI compute. You will craft algorithms for LLM inference on memory-resident hardware, balancing numerical methods with hardware realities and validating in simulation and on silicon.

You will collaborate with hardware, architecture, and software teams, translate model workloads into hardware constraints, and iterate rapidly as the platform evolves from simulation to silicon, with a focus on

Qualifications

• Demonstrated ability to design and implement algorithms for ML or AI workloads.
• Strong foundation in numerical methods and probabilistic reasoning.
• Experience bridging software and hardware design, particularly for specialized accelerators.
• Familiarity with memory/computation trade-offs and in-memory computing concepts.
• Ability to translate model workloads into hardware constraints and benchmarks.

Responsibilities

• Develop algorithms for LLM inference on in-memory hardware.
• Collaborate with hardware/architecture teams to shape hardware capabilities.
• Design numerical methods leveraging analog dynamics and noise.
• Build evaluation frameworks and benchmarks for hardware/simulation.
• Translate workload insights into hardware design constraints and opportunities.
• Prototype rapidly as hardware evolves from simulation to silicon.
• Optimize performance at both algorithmic and gate levels.

Skills

Python proficiency

Systems programming

Collaborative teamwork

Problem solving

Education

PhD in ML / applied mathematics / physics / EE

Tools

Python

C/C++

Job description

Normal Computing is seeking a skilled algorithm and hardware co-design engineer to advance in-memory AI compute. You will craft algorithms for LLM inference on memory-resident hardware, balancing numerical methods with hardware realities and validating in simulation and on silicon.

You will collaborate with hardware, architecture, and software teams, translate model workloads into hardware constraints, and iterate rapidly as the platform evolves from simulation to silicon, with a focus on