LEAD/MANAGER AI OPTIMIZATIONS
Description
Esperanto is creating a new generation of Artificial Intelligence processing solutions delivering unmatched energy efficiency (performance/watt), scalability, and flexibility. The historical limitations on many-core designs have always been complexity and power consumption, but it is now possible to create designs with over a thousand processor cores on one piece of silicon. Esperanto’s no-compromise solution breaks through previous barriers in delivering a massively parallel and flexible processor that combines high performance with ultra-low power consumption. Our high-performance ET-MaxionTM core is designed to deliver the best single-thread RISC-V performance. Our ET-MinionTM core compute array is designed for energy efficiency, delivering TeraFlops and TeraOps of computing.
About the Role
We are hiring a hands-on Software Developer Lead with experience in Generative AI that will develop neural network optimizations for Machine Learning accelerators with thousands of cores. In this role you will be responsible for defining, implementing, and validating algorithmic and compiler solutions to deliver unmatched energy efficiency at inference stage for our HW accelerators. You will be able to explore novel generative AI applications and adapt them to get maximum benefit of our HW architecture. As part of our dynamic SW team, you will be able to actively participate in different components of the SW stack, such as the Compiler, the Runtime, the ML Framework, and the Firmware.
What you will do
- Define, plan, and lead the work on software optimizations for transformer models
- Develop software-driven optimizations targeting a novel RISC-V architecture
- Implement algorithmic and compiler solutions for our high-performance HW accelerators
- Propose solutions to enhance the efficiency of Machine Learning applications
- Integrate different inter-operator optimizations in the whole SW stack
- Analyze and characterize the efficiency of the SW optimizations
- Interact with functional and performance simulators
Minimum Qualifications
- MS in Computer Science or a related technical field
- Good understanding of object-oriented programming
- Knowledge of parallel programming
- Knowledge of algorithmic complexity, data structures, and graph algorithms
- Seasoned experience in C++ and Python
- Working experience in Linux distributions
- Business fluent English
- This is a hybrid position located in Barcelona
Valuable Qualifications
- Knowledge of recent C++ extensions (C++14, C++17, C++20)
- Experience in Pytorch and ONNX Runtime
- Knowledge of kernel vectorization
- Experience with basic GitLab-CI or related CI pipeline flows
- Knowledge of unittest libraries (gtest, pytest)
- Experience in ML compiler development
- Knowledge in high-performance computing
- Knowledge of processor and SOC architecture
- Knowledge of AI/ML algorithms and Neural Networks (CNNs, recommending systems, transformers, LLMs, VLMs, stable diffusion, whisper)