Computer Architect

Acceler8 Talent is looking for an experienced Computer Architect Engineer to join a high-performing San Francisco-based team, helping them create best-in-class silicon for high-performance and sustainable Generative AI. If you’re looking to make an impact in this exciting age of technological advancement, this opportunity may be for you!

Founded by engineers who have been instrumental in the industry’s most successful semiconductor and AI products, this company is pushing the boundaries for what LLMs can accomplish. With a team comprised of industry veterans, talent and passion are the foundational pillars behind their success.

As a Computer Architect, you would drive the design of scalable, high-performance compute architectures and connectivity topologies tailored for modern AI workloads. You would design advanced memory subsystems, including low-latency, high-throughput interconnects and cache hierarchies, and work closely with hardware, firmware, and software teams to ensure specifications meet cross-functional requirements.

This role might be ideal for you if you have:

• Concept-to-silicon ownership in driving silicon architecture for key features at a subsystems and/or chip-level from an architecture specification to production silicon
• Comfort with some of: SystemVerilog, Python, C/C++, Rust, or Bluespec and similar scripting and programming languages for chip architecture and silicon design
• Successful delivery of executable specifications as models to describe hardware architecture
• Track record of delivering innovative features and techniques in chips and systems related to ML, compute, networking, storage and related areas in shipping products
• Experience with debugging architectural issues at silicon and system level with partner teams across: design, validation, system engineering, software and field support
• Experience in writing firmware and software code to test architectural features at product level is a big plus

Experience with SystemC, Python, C/C++ and the ability to write production-ready, annotated code