Candidates should possess a Bachelor’s or Master’s degree in Electrical Engineering or Computer Engineering, along with at least 3 years of proven experience in formal verification, specifically Formal Property Verification. They must have hands-on experience with formal verification tools like Cadence JasperGold or Synopsys VC Formal, and demonstrate expertise in SystemVerilog Assertions (SVAs). A solid understanding of RTL design using Verilog or VHDL, combined with knowledge of digital design principles, is also required. Furthermore, candidates should have a decent understanding of compute architecture, including CPU, GPU, or specialized AI hardware accelerators, and the ability to verify performance, timing, and functional properties within these architectures. Proficiency in scripting languages, particularly Python, for automation and tool integration is essential.
The Senior/Staff Formal Verification Engineer will drive formal verification efforts to ensure the correctness and performance of Custom-AI hardware designs at both IP and Sub-system level. They will develop and execute verification plans, including creating formal property test plans and providing final sign-off. Collaboration with hardware designers and architects is crucial to define verification requirements and ensure alignment with system-level goals. The role involves analyzing and debugging formal verification results, providing feedback to design teams, and contributing to the continuous improvement of verification methodologies and processes. Staying current with advancements in formal methods and AI hardware verification is also a key responsibility, alongside applying new techniques to enhance verification strategies.
Enhances AI inference with at-memory computing
Untether AI enhances the speed and efficiency of AI inference workloads using at-memory computing. This method places the compute element next to memory cells, which boosts compute density and accelerates AI inference for various neural networks, such as those used in vision, natural language processing, and recommendation systems. The company targets businesses that rely on AI technologies and need high-performance computing for inference tasks. Their products, including the runAI200® devices and tsunAImi® accelerator cards, are designed to deliver exceptional performance, with the tsunAImi® card offering over 2 PetaOps. This allows businesses to optimize their AI workloads while maintaining a compact PCI-Express form factor. Untether AI's goal is to provide efficient and cost-effective solutions for companies looking to enhance their AI applications.