Unconventional, Inc.

Role Description As a Member of Technical Staff, System Modeling (Dynamic Systems Simulation), you will be part of a hands-on R&D team building simulation frameworks that enable testing and rapid iteration across all layers of unconventional physics-based computing systems for machine learning workloads. “Extreme co-design” is our guiding principle. System Modeling is a multi-disciplinary effort, and the team we’re building reflects that. The role involves: - Development of physics-based system models - GPU-accelerated ML system simulations - Cross-layer system integration You don’t need to be an expert in all of these, but you have to be very strong in at least one, and solid in the rest. You will be responsible for developing high-performance PyTorch components that model complex, time-varying dynamic systems. Your work will directly enable next-generation AI architectures, requiring a holistic approach involving everything from high-level neural network design down to the fundamental differential equations that govern system behavior. Qualifications - MS/PhD in a quantitative field (AI/ML, Computer Science, Physics, Electrical Engineering, Applied Math), or BS with substantial, clear evidence of equivalent research/engineering depth. - Dynamical systems simulation knowledge - Advanced Neural Modeling (PyTorch): - Deep proficiency in PyTorch, specifically in building custom autograd functions and integrating numerical solvers (e.g., Neural ODEs) to represent dynamic processes. - Dynamics & Differential Equations: - A strong theoretical and practical grasp of linear and non-linear dynamics, state-space representations, and solving $dx/dt = f(x, u, t)$ within a machine learning context. - Stochastic Processes & Noise: - Understanding how to model and mitigate noise in real-world systems, including experience with stochastic differential equations (SDEs) or Bayesian filtering. - Modeling & Simulation (M&S): - Proven industry experience building high-fidelity simulations that balance computational efficiency with physical accuracy. - Systems Engineering (Analog/Digital): - Familiarity with hardware-level concepts like circuit dynamics, signal processing, or transfer functions is highly desirable to help ground our digital models in physical reality. - ML and systems fluency: - Solid understanding of modern AI/ML architectures and training/inference workflows. - Strong experience implementing and debugging ML models in PyTorch (preferred) or similar, with practical experience profiling, optimizing, and stabilizing non-trivial large-scale ML systems. Requirements - Strong Python engineering skills: modular design, testing, packaging, CI. - Experience with PyTorch internals: autograd, custom modules, low-level ops; familiarity with torch.compile or similar graph capture/compile flows. - Experience with CUDA, Triton, or other GPU programming approaches (writing custom kernels, understanding memory hierarchy, basic performance tuning). - Comfort with at least some of: JAX, NumPy, TensorFlow, Modal, HPC patterns (MPI, NCCL, distributed training), SciPy. - Demonstrated ability to reason across multiple layers of the stack: algorithm, software, runtime, hardware. - Able to connect model architecture choices to system performance implications: memory bandwidth, communication patterns, latency, energy, and numerical issues. - Experience applying at least some efficiency techniques (quantization, sparsity, pruning, distillation, kernel fusion, etc.). Benefits - A comprehensive package including best-in-class health benefits - 401k matching - Truly unlimited PTO - Complimentary meals when working from our Palo Alto office

Role Description As a Member of Technical Staff, System Modeling (Performance Models), you will be part of a hands-on R&D team building simulation frameworks that enable evaluation and rapid iteration across all layers of unconventional physics-based computing systems for machine learning workloads. “Extreme co-design” is our guiding principle. System Modeling is a multi-disciplinary effort, and the team we’re building reflects that. The role involves: - Development of physics-based system models - GPU-accelerated ML system simulations - Cross-layer system integration You don’t need to be an expert in all of these, but you have to be very strong in at least one, and solid in the rest. Qualifications - MS/PhD in a quantitative field (AI/ML, Computer Science, Physics, Electrical Engineering, Applied Math), or BS with substantial, clear evidence of equivalent research/engineering depth. - Experience with tools and development for power profiling, modeling and simulation for AI workloads. - Deep understanding of spatial architectures and data orchestration mechanisms. - Deep understanding of different dataflow strategies and their tradeoffs, e.g. Weight-Stationary (WS), Output-Stationary (OS), Input-Stationary (IS) and Row-Stationary (RS). - Familiar with (OSS) tools for hardware accelerator design: TimLoop, Accelergy, NeuroSim, CIMLoop, CACTI, etc. - Familiar with different existing systolic array accelerator architectures for AI/ML workloads. - Solid understanding of modern AI/ML architectures and training/inference workflows. - Strong experience implementing and debugging ML models in PyTorch (preferred) or similar, with practical experience profiling, optimizing, and stabilizing non-trivial large-scale ML systems. Requirements - Basic familiarity of analog dynamic systems, including transient responses, nonidealities such as nonlinearity, quantization, random noise, and feedback/stability. - Strong Python engineering skills: modular design, testing, packaging, CI. - Experience with PyTorch internals: autograd, custom modules, low-level ops; familiarity with torch.compile or similar graph capture/compile flows. - Experience with CUDA, Triton, or other GPU programming approaches (writing custom kernels, understanding memory hierarchy, basic performance tuning). - Comfort with at least some of: JAX, NumPy, TensorFlow, Modal, HPC patterns (MPI, NCCL, distributed training), SciPy. - Demonstrated ability to reason across multiple layers of the stack: algorithm, software, runtime, hardware. - Able to connect model architecture choices to system performance implications: memory bandwidth, communication patterns, latency, energy, and numerical issues. - Experience applying at least some efficiency techniques (quantization, sparsity, pruning, distillation, kernel fusion, etc.). - Prior experience building or extending a serious simulation or modeling framework (could be ML systems, physics, circuits, or other technical domains). - Comfort with approximations and tradeoffs: you know when to use a simple model and when you need something closer to the physics. Benefits - A comprehensive package including best-in-class health benefits - 401k matching - Truly unlimited PTO - Complimentary meals when working from our Palo Alto office

Role Description As a Member of Technical Staff, System Modeling (Computation), you will be part of a hands-on R&D team building simulation frameworks that enable evaluation and rapid iteration across all layers of unconventional physics-based computing systems for machine learning workloads. “Extreme co-design” is our guiding principle. System Modeling is a multi-disciplinary effort, and the team we’re building reflects that. The role involves: - Development of physics-based system models. - GPU-accelerated ML system simulations. - Cross-layer system integration. You don’t need to be an expert in all of these, but you have to be very strong in at least one, and solid in the rest. Responsibilities - Building large-scale, GPU-accelerated, high-fidelity differential equation solvers to enable development of ML on analog/unconventional hardware. - Developing physics-based models of device- and system-level behavior in unconventional compute, integrated with PyTorch. - Working with other teams to understand their needs for modeling and simulation to support high-level algorithm development as well as lower-level verification of hardware. Qualifications - MS/PhD in a quantitative field (AI/ML, Computer Science, Physics, Electrical Engineering, Applied Math), or BS with substantial, clear evidence of equivalent research/engineering depth. - Dynamical computation knowledge. - Experience with high-performance, customized GPU kernel development for numerical differential equation solvers (e.g., ODE, SDE). - Experience with DE solving algorithms designed for parallel architectures. - Experience with building effective neural network surrogate models for ODE/SDE/DDE solvers and functions. - Solid understanding of modern AI/ML architectures and training/inference workflows. - Strong experience implementing and debugging ML models in PyTorch (preferred). Preferred Qualifications - Basic familiarity of analog dynamic systems, including transient responses and nonidealities. - Strong Python engineering skills: modular design, testing, packaging, CI. - Experience with PyTorch internals and CUDA, Triton, or other GPU programming approaches. - Comfort with at least some of: JAX, NumPy, TensorFlow, Modal, HPC patterns. - Demonstrated ability to reason across multiple layers of the stack: algorithm, software, runtime, hardware. - Experience applying efficiency techniques (quantization, sparsity, pruning, distillation, kernel fusion, etc.). - Prior experience building or extending a serious simulation or modeling framework. - Comfort with approximations and tradeoffs. Benefits - A comprehensive package including best-in-class health benefits. - 401k matching. - Truly unlimited PTO. - Complimentary meals when working from our Palo Alto office.

Role Description As a Member of Technical Staff, you will be a foundational member of our small, multi-disciplinary R&D team. We are looking for 'first principles' thinkers who are excited to tackle the hardest, most ambiguous technical challenges at the intersection of AI, physics, and computer architecture. You will be responsible for driving invention, prototyping, and validation of the core components of our novel computing platform. Your work will be fluid and could span from theoretical modeling and simulation to algorithm development, hardware/software co-design, or experimental validation in collaboration with other team members. We're hiring exceptional problem-solvers who can navigate deep uncertainty and help chart our technical roadmap. Qualifications - Exceptional technical ability in a quantitative field (e.g., Physics, Computer Science, Electrical Engineering, Applied Math, or a related discipline). An MS/PhD or equivalent research/project experience is strongly preferred. - A "0-to-1" mindset. You have a demonstrated history of tackling complex, ambiguous R&D problems, often from a blank slate. - Deep curiosity. You are comfortable diving into new domains, whether it's semiconductor physics, machine learning theory, or systems-level design. - A creative and unconventional approach to problem-solving. Requirements - Core competences in the analysis of nonlinear dynamical systems (ODEs, PDEs, SDEs), ideally with experience analyzing the stability, noise robustness, and capacity of such systems. - The ability to leverage analytic insights to build practical tools – metrics, algorithmic optimizations, and automated analyses – that can be used to study dynamical systems. - Strong command of Python and expertise with using numeric computing and visualization libraries, such as numpy, scipy, and matplotlib. - Familiarity with dynamics-based ML model architectures, such as diffusion models and energy-based models, and general experience with ML model training flows. - Experience with using high-level ML model frameworks, such as PyTorch and JAX. Bonus Points (Nice to Have) - Compute model technical staff may focus primarily on the above skillsets, or may be cross-disciplinary with hardware or AI/ML algorithms expertise. - Excellent ability to translate complex technical concepts for diverse teams. - You will act as a translator, discussing algorithmic/model trade-offs with ML/AI teams and eliciting hardware constraints and features from hardware engineering teams. Benefits - A comprehensive package including best-in-class health benefits. - 401k matching. - Truly unlimited PTO. - Complimentary meals in our Palo Alto office.

Role Description You will be a key contributor to our training ecosystem. Your goal is to build the next-generation ML model training platform tailored for a world where compute is no longer constrained by the digital abstraction. You will co-design and implement training systems alongside novel AI models and hardware platforms that push the boundaries of physics-based compute. What You’ll Do - The Model Architectures: Build and maintain highly optimized, model-specific training stacks specifically tuned for state-of-the-art generative vision, language, and world models. - The Training Infrastructure: Design and scale multi-node distributed training systems, implementing elastic sharding and robust data streaming pipelines for fast, large-scale iteration. Implement robust model checkpointing and recovery mechanisms. - Optimization & Benchmarking: Develop and optimize kernels using low-level programming models like CUDA and Triton. Design rigorous benchmarking suites to track Model Flops Utilization (MFU), memory bandwidth, and convergence stability. - Cross-Functional Collaboration: Act as a translator, discussing algorithmic trade-offs with theorists and converting model requirements into concrete specifications for infrastructure and hardware engineering teams. Qualifications - Education: An MS/PhD or equivalent research/project experience in a quantitative field such as AI/Machine Learning, Computer Science, Physics, Electrical Engineering, or Applied Math. - Experience: Veteran of the modern ML software stack. Demonstrated ability to map state-of-the-art AI model architectures (e.g., transformers, Mixture of Experts, diffusion models) to system performance implications. Deep expertise in how models are partitioned across a cluster, with mastery of communication primitives and parallelism strategies. - Software Development: Proven track record of implementing, debugging, and maintaining production-grade training frameworks—such as Megatron-LM, DeepSpeed, Ray, PyTorch Lightning—turning raw compute into a reliable model-building factory. Preferred Qualifications (Nice to Have) - Unconventional Co-Design: A forward-looking perspective on co-designing algorithms for unconventional computing paradigms that map closely to the physics of underlying systems. Benefits - The Mission: Redefine computing for the next 50 years by solving the fundamental energy limitation of AI at a global scale. - The Impact: Shape the company's future as a foundational team member. Enjoy massive ownership and an outsized opportunity to drive change. - The Perks: A comprehensive package including best-in-class health benefits, 401k matching, truly unlimited PTO, and complimentary meals in our Palo Alto office.

Role Description As a Member of Technical Staff, Language & Reasoning Models, you will drive the development of foundational language and reasoning models that fundamentally leverage the dynamics of our novel silicon. Your goal is to map the behaviors of modern language models directly onto the physics of our hardware. You will sit at the intersection of NLP/reasoning research and hardware codesign, proving that high-fidelity, large-scale language understanding and generation can be achieved natively on an unconventional computing substrate. What You'll Do - Model Development: Design, train, and scale next-generation language and reasoning architectures (such as transformers, state space models, diffusion/flow models, and deep equilibrium models) specifically tailored for unconventional compute. - Physics-Informed Architecture: Rethink standard sequence modeling to exploit the continuous-time dynamics of silicon, moving away from layers of inefficient digital abstraction. - Evaluation & Scaling: Establish the training recipes, loss functions, and evaluation metrics needed to reach the frontier of language comprehension, logical reasoning, and generation speed while maintaining the massive energy efficiency of our platform. - Extreme Codesign: Collaborate with hardware designers and theorists, and system builders to co-design the model architecture alongside the underlying physical compute primitives. Qualifications - Education: An MS/PhD or equivalent research/project experience in a quantitative field such as AI/Machine Learning, Computer Science, Physics, Electrical Engineering, or Applied Math. - Experience: Deep, hands-on expertise in the theory, architecture, and training of modern foundation models (transformers, SSMs, text diffusion/flow, etc.). - Systems Fluency: Hands-on, battle-tested experience dealing with model scaling. You have successfully designed and executed full-scale, distributed training runs for large language or reasoning models, managing the complexities of massive compute clusters. - Software Development: You are fluent in modern deep learning frameworks (PyTorch or JAX) and have a proven track record of writing clean, scalable training code for large language models. Requirements - Unconventional Experience: As a bonus, you may have experience working with hardware-in-the-loop training, mixed-signal hardware, quantization, or physics-informed neural networks. Benefits - A comprehensive package including best-in-class health benefits. - 401k matching. - Truly unlimited PTO. - Complimentary meals in our Palo Alto office.

Role Description - Define, architect, and implement comprehensive verification plans and environments for IP blocks and full chip from specification to tape-out. - Help build verification infrastructure from ground zero to flexible and nimble environments, able to move quickly and scale for growth. - Take full ownership of the verification lifecycle, including the development and maintenance of scalable testbenches, sophisticated test cases, and robust verification environments. - Deploy industry-standard methodologies (UVM/SystemVerilog) to conduct constrained-random stimulus generation, coverage-driven verification, and assertions. - Collaborate closely with digital and analog design teams, as well as software and systems teams, to define verification requirements, debug complex hardware/software interactions, and ensure design integrity. - Perform exhaustive analysis of functional and code coverage. Debug failures at the RTL and gate level to ensure 100% verification closure and high-quality silicon. Qualifications - Bachelor’s degree in Electrical Engineering, Computer Engineering, Computer Science, or a related field, or equivalent practical experience. - Minimum of 8 years of hands-on experience with advanced verification methodologies and languages, specifically UVM and SystemVerilog. - Proven track record of developing and maintaining complex verification testbenches and automated test environments. - Proficiency in scripting languages (e.g., Python, Shell, or Bash) for automation of verification flows and result analysis. - Strong familiarity with general-purpose operating systems such as Linux, and industry standard tools from companies such as Cadence and Synopsys. Requirements - Master's degree or Ph.D. in Electrical Engineering or Computer Science, with a specific emphasis on computer architecture (preferred). - Expert-level knowledge of UVM architecture and high-level behavioral modeling (preferred). - Experience with modern AI-driven code generation tools (preferred). Benefits - A comprehensive package including best-in-class health benefits. - 401k matching. - Truly unlimited PTO. - Complimentary meals in our Palo Alto office.

Role Description As a Junior Member of Technical Staff, System Modeling, you will work closely with senior engineers to contribute to the development of our multi-disciplinary simulation frameworks. You will assist the hands-on R&D team in building simulation environments that enable rapid iteration and testing across all layers of our unconventional physics-based computing systems for machine learning workloads. Your work will focus on integrating physics-based models, developing GPU-accelerated simulations, and supporting the cross-layer system integration necessary for "Extreme co-design". Key Responsibilities - Contribute to the implementation and optimization of GPU-accelerated simulators for ML on analog/unconventional hardware, focusing on specific modules and features within PyTorch. - Assist in integrating physics-based device and system models into the PyTorch simulation environment to help expose early algorithm–hardware tradeoffs and enable cross-layer optimization. - Support the maintenance and extension of the unified end-to-end simulation environment, helping to link theory, algorithms, and device models, and ensuring alignment between high-level and near-physical simulators. - Help implement and adhere to robust experiment tracking protocols to ensure simulation results, configurations, and non-idealities are reproducible and auditable. - Collaborate with Algorithms and Hardware teams to gather requirements and ensure the modeling environment meets their needs for high-level algorithm development and lower-level hardware verification. Qualifications - A BS, MS, or PhD in Computer Science, Electrical Engineering, or a related technical field. - Deep understanding of computer architecture and operating systems. - Strong skills in C++ and Python; comfortable writing performance-critical code. - Basic familiarity with the internals of deep learning frameworks (e.g., how a PyTorch graph is executed) and common model architectures. - A solid grasp of linear algebra and calculus, essential for understanding both neural dynamics and hardware optimizations. - Enjoy digging into "why" things work (or don't) and challenge conventional software "best practices". Bonus Points - Experience with compilers (LLVM, MLIR) or domain-specific languages like Triton. - Exposure to GPU programming (CUDA) or other hardware accelerators. - Prior research or internship experience in high-performance computing (HPC) or neuromorphic systems. - Contributions to open-source AI or systems software projects. Benefits - Learn directly from the architects who built the modern AI stack at companies like Intel, Databricks, and NVIDIA. - You won't be a small cog in a giant machine; you will be helping build the machine itself. - Work on challenges that don't have a StackOverflow answer—you’ll be defining the future of AI compute. - Significant equity and competitive salary at a well-funded, high-growth startup.