Software Engineer

Efficient Computer is rethinking computing from the ground up to solve one of AI's biggest challenges: energy. Our Fabric architecture delivers a 10–100x improvement in energy efficiency for general-purpose computation, including AI, while supporting familiar programming languages and software frameworks—unlocking new possibilities for autonomous machines, intelligent infrastructure, wearables, and space systems. With our first processor, Electron E1, now shipping in volume and more than $97 million in Series B financing recently announced, we're accelerating customer deployments and scaling our architecture to data center performance. Join us to build the hardware and software that bring intelligence wherever it's needed and help shape the next generation of computing.

Efficient Computer is building the next generation of energy efficient, general-purpose processors designed around the Fabric architecture, a novel spatial dataflow design spun out from decades of fundamental SOC research at CMU.

We are looking for a Senior Staff Linux Runtime Software Engineer to own the user-mode components that interface with the Fabric and support the effcc compiler.  This includes owning the user-mode driver, interfacing with our kernel mode driver,  as well as defining the HAL interface exposed by the Fabric ABI.  You will be working within pre-silicon environments, involving QEMU, functional simulation, emulation, and FPGAs to ensure applications run on day one.   

This is a key role for the System Software team.   As such, you will work closely with the compiler, kernel driver, architecture, micro-architecture, and verification teams during all phases of development.   

Key Responsibilities

Own the Runtime Library and User-Mode Driver

• Design, implement, and maintain the user-mode runtime driver for the Fabric accelerator.  This is a critical shared library which sits between the compiler and the Fabric ABI.  This library implements context management, program loading, resource management, command queue construction and submission, descriptor allocation and setup, streaming, completion queue/event synchronization, error handling,  logging, memory allocation, as well as general library support.

• Design and implement all user-mode interfaces between the compiler, the Fabric ABI, and the kernel mode driver.

• Define the in-memory representations generated by the compiler and consumed by the Fabric.

• Design and implement the underlying runtime library support and interfaces needed by 3rd party ML framework backends used by PyTorch, ONNX, and/or TensorFlow.

• Lead the user-mode efforts to bring up on first silicon.  This includes being comfortable with functional simulation, emulation and FPGA environments.