Job description

ABOUT NEUROPHOS The demand for new data centers and AI compute is rapidly outpacing the planet's energy capacity. Digital solutions are hitting a power wall as we approach the physical limits of traditional silicon. Conquering this bottleneck means rethinking the fundamental architecture of inference compute. The industry's current path can't meet the need, so we're taking a different approach. Instead of traditional electronic circuits, we use silicon photonics and an active, programmable metasurface to perform matrix multiplications at the speed of light. Our optical cells are 10,000x smaller than traditional photonic components, enabling unprecedented density. By using photonics instead of electricity, our chips become more efficient as they scale. This architecture will deliver up to 100 times the energy efficiency of existing solutions while significantly improving performance for large-scale AI inference. We’ve assembled a world-class team of industry veterans and recently raised a $110M Series A neurophos.com led by Gates Frontier. Participants include M12 (Microsoft’s Venture Fund), Carbon Direct Capital, Aramco Ventures, Bosch Ventures, Tectonic Ventures, Space Capital, and others. Join us and shape the future of computing! Location: Austin, TX or Sunnyvale, CA. Full-time onsite position. Reports To: Mathematics and Quantization Lead FLSA Status: Exempt POSITION OVERVIEW We are seeking an experienced machine learning scientist to develop advanced post-training quantization methods for large language models (LLMs), diffusion models, and other ML applications for our revolutionary optical inference engines. This role is critical to demonstrating the full potential of our metamaterial-based optical processing units (OPUs) by adapting state-of-the-art AI models to leverage our ultra-high-throughput, low-precision compute architecture. The ideal candidate will bridge the gap between cutting-edge ML research and novel hardware capabilities, ensuring customers can seamlessly deploy their AI workloads on Neurophos hardware. KEY RESPONSIBILITIES - Develop and execute hardware-aware post-training methods for full model quantization. - Investigate preconditioning and formulate quantization as non-convex, discrete, constrained, or second-order optimization and develop practical solutions. - Contribute to refining Neurophos's quantization strategy. - Design controlled numerical experiments to understand potential improvements and secondary effects due to analog processing hardware. - Build research-quality implementations and reproducible experiment harnesses for testing candidate methods. - Adapt models from open-source repositories and customer private models. - Work with models in various formats, including PyTorch, Triton, JAX, and emerging frameworks. - Design and execute re-quantization, retraining, and other model adaptation techniques to minimize accuracy loss during precision reduction. - Optimize GEMM operations for high-throughput execution. - Collaborate with hardware, software, and architecture teams to co-optimize model architectures for optical compute characteristics. - Publish research papers on novel optimization techniques and methodologies, with appropriate IP protection. QUALIFICATIONS - PhD, or equivalent research experience, in machine learning, applied mathematics, optimization, numerical analysis, computer science, or a closely related field. - 5+ years of experience in machine learning, with at least 3 years focused on model optimization and deployment. - Research or advanced engineering experience in neural network quantization, model compression, numerical optimization, or efficient inference. - Strong knowledge of numerical linear algebra, including matrix factorizations, conditioning, covariance estimation, and iterative methods. - Experience with one or more of non-convex optimization, discrete optimization, manifold optimization, second-order methods, or constrained optimization. - Strong proficiency in PyTorch and familiarity with other ML frameworks, including JAX, Triton, and TensorFlow. - Hands-on experience with transformer architectures, LLMs, and diffusion models. - Experience designing controlled numerical experiments and distinguishing algorithmic improvements from calibration or benchmark artifacts. - Strong written communication and research collaboration skills. PREFERRED SKILLS - Experience with low-precision inference optimization (INT8, FP8, or lower). - Background in analog or optical computing architectures. - Knowledge of in-memory computing paradigms and matrix-vector multiplication acceleration. - Knowledge of randomized numerical linear algebra, sketching, or structured transforms. - Publications in quantization, optimization, numerical linear algebra, model compression, or efficient ML. - Experience with vector quantization, lattice methods, learned codebooks, or rate-distortion ideas. - Experience with large-scale batch inference optimization. - Familiarity with prefill versus decode optimization strategies in LLM inference. - Experience conducting experiments on models large enough to expose scaling and generalization problems. WHAT WE OFFER This is an opportunity to play a pivotal role in an innovative startup redefining the future of AI hardware. Work on game-changing technology at the intersection of photonics and AI as part of a collaborative, brilliant team. You’ll contribute to a platform that redefines computational performance and accelerates the future of artificial intelligence. Come help us bring this transformative technology to the world. BENEFITS Join a team that invests in your future and your well-being. At Neurophos, we offer: - 100% coverage of base health plan premiums for you and your dependents, plus HSA contributions. - Unlimited PTO. No rigid vacation banks, just a focus on delivery. - 401(k) matching and stock option opportunities to ensure our success is your success. - Full suite of voluntary benefits, including Dental, Vision, Life, Hospital, Critical Illness, and Accident insurance. - Personalized Benefits. Choose the plans that fit your life and take the cash back for those that don’t.

Requirements

Must-haves

The ideal candidate will bridge the gap between cutting-edge ML research and novel hardware capabilities, ensuring customers can seamlessly deploy their AI workloads on Neurophos hardware.
Develop and execute hardware-aware post-training methods for full model quantization.
Investigate preconditioning and formulate quantization as non-convex, discrete, constrained, or second-order optimization and develop practical solutions.
Contribute to refining Neurophos's quantization strategy.
Design controlled numerical experiments to understand potential improvements and secondary effects due to analog processing hardware.
Build research-quality implementations and reproducible experiment harnesses for testing candidate methods.
Adapt models from open-source repositories and customer private models.
Work with models in various formats, including PyTorch, Triton, JAX, and emerging frameworks.
Design and execute re-quantization, retraining, and other model adaptation techniques to minimize accuracy loss during precision reduction.
Optimize GEMM operations for high-throughput execution.
Collaborate with hardware, software, and architecture teams to co-optimize model architectures for optical compute characteristics.
Publish research papers on novel optimization techniques and methodologies, with appropriate IP protection.
PhD, or equivalent research experience, in machine learning, applied mathematics, optimization, numerical analysis, computer science, or a closely related field.
5+ years of experience in machine learning, with at least 3 years focused on model optimization and deployment.
Research or advanced engineering experience in neural network quantization, model compression, numerical optimization, or efficient inference.

Nice-to-haves

Experience with low-precision inference optimization (INT8, FP8, or lower).
Background in analog or optical computing architectures.
Knowledge of in-memory computing paradigms and matrix-vector multiplication acceleration.
Knowledge of randomized numerical linear algebra, sketching, or structured transforms.
Publications in quantization, optimization, numerical linear algebra, model compression, or efficient ML.
Experience with vector quantization, lattice methods, learned codebooks, or rate-distortion ideas.
Experience with large-scale batch inference optimization.
Familiarity with prefill versus decode optimization strategies in LLM inference.
Experience conducting experiments on models large enough to expose scaling and generalization problems.

Application prep

Use the role’s language and application details while tailoring your materials.

Keywords to mirror

aimachine learningmlllm

Platform: ashby
Effort: quick

Benefits

Benefits detected in the listing.

Dental

Vision

401(K)

Pto