About Unconventional AI

Unconventional AI is rethinking the foundations of a computer to optimize energy efficiency for AI. Founded by experts in AI systems, analog circuits, computing theory, and neuroscience, we are bringing biology-scale efficiency to artificial intelligence. We design and prototype a new computational substrate by building analog silicon chips that emulate the non-linear dynamics of biological neurons, aiming for significantly higher performance-per-watt than traditional GPUs.

The Role

As a Hardware Systems Engineer, you will play a pivotal role in bringing our groundbreaking analog AI hardware to life. You will lead the transition from "taped-out" silicon to a fully functional, characterized system. You will be responsible for the first power-on of our chips, driving deep-dive debugging across hardware and software boundaries, and building the automated infrastructure required to validate our "biology-scale" efficiency.

Responsibilities

• Silicon Bring-up & Debug: Drive the early bring-up of first silicon, including boot, initialization, and root-cause analysis of complex hardware/firmware issues.
• Automated Infrastructure: Develop and maintain Python-based automation frameworks and test suites for high-volume system-level testing and data collection.
• Lab Leadership: Take ownership of the lab environment, test equipment and designing custom test fixtures or test boards.
• Performance Profiling: Conduct power and thermal characterization to correlate silicon performance with architectural power models and AI workload benchmarks.
• Validation & Characterization: Execute comprehensive validation plans for high-speed interfaces, digital and analog compute blocks.
• Cross-Functional Collaboration: Partner closely with RTL Design, Analog, and Software teams to develop firmware workarounds for silicon errata and influence the "design for debug" (DFD) features of future generations.

Qualifications

• Education: BS/MS in Electrical Engineering, Computer Engineering, or Physics.
• Experience: 5+ years of hands-on experience in pre and post-silicon validation, bring-up, and characterization of high-performance SoCs or AI accelerators.
• Technical Depth: Deep understanding of SOC architecture, FPGA-based emulation, analog circuits, high-speed serial protocols (PCIe, USB) and memory subsystems (LPDDR5, HBM3).
• Lab Expertise: Proficient with high-speed lab instrumentation, including real-time oscilloscopes, protocol analyzers, and JTAG debuggers.
• Programming: Strong scripting skills in Python for test automation and C/C++ for low-level embedded firmware/driver development.
• Hardware Fundamentals: Solid understanding of EE fundamentals such as digital/analog design, power management, clocking architectures, and signal integrity.
• Mindset: A self-starter who thrives in a "startup-vitality" environment and is comfortable navigating the ambiguity of first-of-their-kind analog AI architectures.

Why Join Us?

At Unconventional AI, we foster a culture of "humble expertise" and bold execution. You will work on a breakthrough AI platform that moves beyond the constraints of the GPU, collaborating with a team that has previously led industry-defining projects at Intel, Nervana, and MosaicML. We offer competitive salaries, significant equity, and the opportunity to help shape the energy-efficient future of the AI landscape. Join us in building the world’s most efficient computational substrate.