Various

Company Description -

MemWize is a VC-backed, stealth-mode startup building rack-level AI inference system. We are seeking highly skilled and motivated engineers to join us in building a uniquely differentiated SoC, system and developing/extending open-source SW to serve leading-edge models with extreme efficiency.

Team members we are looking for-

Looking to fill various positions for senior, experienced SW and HW engineers. Currently seeking across following key roles. Currently we are only looking to fill full-time, on-site positions at our offices in Santa Clara, CA.

• CUDA developers to build and optimize high-performance GPU kernels for next-generation AI systems.
• Machine learning engineers to drive performance improvements in inference frameworks (vLLM, SGLang, PyTorch) and develop advanced cluster scheduling algorithms to optimize throughput and latency
• Performance modeling engineers to develop functional models and performance models at multiple levels of abstraction that guide architecture decisions and accelerate hardware–software co-design
• Machine learning compiler developers to develop new ML compiler capabilities in a modern ML compiler stack from PyTorch through Triton down to CUDA/ROCm or machine IR on leading edge GPUs and new accelerators
• SoC Architects to define full-chip or significant parts of the SoC, spanning across high-end connectivity options, memory interfaces, data movement architecture and application processor complex and control plane definition
• Micro-architect to define the data movement algorithms and pathways between all interfaces and compute blocks on an SoC in a leading-edge process technology
• Micro-architect and RTL developers to define and implement compute blocks with special focus on low power
• SoC RTL leads and developers to define and implement various blocks, working closely with architects and micro-architects