Senior Staff Software Engineer, Linux Runtime & Drivers

Location: San Francisco, Bay Area or Pittsburgh, PA, United States

Company

hirify.global develops energy-efficient general-purpose processors based on its Fabric spatial dataflow architecture for AI, autonomous machines, intelligent infrastructure, wearables, and space systems.

What you will do

• Own the user-mode runtime driver and shared runtime library for the Fabric accelerator.
• Design context management, program loading, resource management, command queues, descriptors, streaming, synchronization, memory allocation, logging, and error handling.
• Define interfaces between the compiler, Fabric ABI, and kernel-mode driver, including in-memory representations and public runtime APIs.
• Build runtime support for PyTorch, ONNX, and TensorFlow backends, including Python bindings and C/C++ interfaces.
• Lead user-mode software bring-up across functional simulation, QEMU, emulation, FPGA environments, and first silicon.
• Establish unit, conformance, regression, performance, and hardware-in-the-loop test coverage.

Requirements

• Bachelor’s degree in Electrical or Computer Engineering, Computer Science, or equivalent experience; a master’s degree is a plus.
• 7+ years of systems software experience, including hands-on Linux user-mode driver and language runtime library development for accelerators.
• Production experience building accelerator runtime stacks, memory models, allocators, zero-copy data paths, and compiler-to-runtime interfaces.
• Experience designing stable public APIs and kernel/user-space interfaces covering ABI versioning, synchronization, memory sharing, error handling, thread safety, and documentation.
• Experience with QEMU, emulation, or simulation in pre-silicon development, plus strong debugging, tracing, profiling, sanitizer, and race-diagnosis skills.
• Expert C and C++, strong Python, working knowledge of ARM/AArch64 Linux and the Linux driver model, and experience with AI-assisted development tools.

Nice to have

• Experience with CUDA Runtime, ROCm, HIP, ROCr, HSA, Tenstorrent Runtime, oneAPI Level Zero, OpenCL, or another accelerator runtime stack.
• Familiarity with LLVM, MLIR, dataflow or spatial architectures, and compiler targeting for accelerators.
• Experience with first-generation silicon products, FPGA prototyping, or Palladium, Zebu, Veloce, or similar hardware emulation platforms.

Culture & Benefits

• Ownership and opportunities to make an impact while building new computing hardware and software.
• Competitive pay and equity.
• 401(k) match and company-paid benefits.
• Paid parental leave and flexibility.
• Opportunities for professional growth as the company scales.