CTO / Co-Founder "Invent the chip for AI on the Edge"

Invent with us the chip, which fulfill the requirements of "AI on the edge". Big LLM models slow down due to many reasons:high energy consumption latencylimitations for bandwidth for communication and memorysecuritynot high enough precision (esp. in health)big privacy concernsask for the full excel list or Claude :-)But AI can do good! And there are so many fields not covered; just think about AI on the edge, focused medium sized models trained with experts controlling (esp health), (Large) Picture Models (LPMs) and so much more. We build solutions for AI on the edge, LPMs and more. With that we understand the requirements for the chip better than most. And we recognized that none of the existing chips is designed for real AI on the edge. And many others recognize this now too. And ask to build and form a consortium respectively Joint Venture. To achieve this, we need a visionary SoC Architect who understands what true high-throughput computing based on neural networks looks like. We are specifically looking for engineering pioneers who have helped shape the industry’s most powerful GPU and mixed-signal compute architectures. If you know how to build architectures that push the absolute limits of bandwidth, parallel processing, and power efficiency, your next challenge is here.What You Will Do: Architect the Disruptor: Define the next-generation microarchitecture for an AI on the edge chip, explicitly targeting performance metrics that surpass the current market leader.Domain Integration: Lead the architectural definition and integration of high-performance CPU clusters, ultra-scalable GPU cores, custom Neural Processing Units (NPUs), and advanced Image Signal Processors (ISPs).Heterogeneous Compute & Memory: Design high-bandwidth memory subsystems (LPDDR5x/HBM) and low-latency interconnects capable of handling simultaneous heavy compute workloads (e.g., AAA gaming, multi-camera processing, and real-time LLMs).Automotive Functional Safety: Partner with safety teams to embed ISO 26262 compliance (ASIL-B/D) directly into the silicon fabric without sacrificing compute throughput.Hardware-Software Co-Design: Collaborate with software engineering teams to optimize silicon for hypervisors, real-time operating systems (QNX/VxWorks), Automotive Grade Linux, and modern AI frameworks.Vendor & Foundry Steering: Drive technical requirements with Tier-1 IP vendors (ARM, Synopsys, etc.) and leading semiconductor foundries for advanced node fabrication (3nm/2nm).What We Are Looking For (Profile Requirements)The Innovation Leader Pedigree: You have a proven track record (5+ years) at a Leading Innovator designing, architecting, or verifying high-performance SoCs, GPUs, or Tegra/Drive platform processors.Deep Architectural Expertise: Extensive experience in RTL design, IP integration, clock/power management architectures, and top-level SoC fabric design.Mixed-Workload Mastery: You understand how to balance the heavy parallel processing demands of modern graphical user interfaces (3D/Unreal Engine) with concurrent neural network execution.Automotive Know-How (Plus but not required): Familiarity with automotive requirements like functional safety (ISO 26262), hardware security modules (HSM), or low-latency camera-to-display pipelines. If you have only worked on data center or gaming silicon but want to apply it to automotive, we want to talk to you.Educational Background: A Bachelor’s, Master’s, or PhD in Electrical Engineering, Computer Engineering, or a highly related technical field.Why Join Us?Pure Impact: Break a global de facto monopoly. Your work will directly define the user experience in millions of next-generation products and solutions.Zero Legacy Code: Build a clean-sheet architecture using the absolute latest advanced node technologies.Elite Peers: Work alongside a highly dense team of industry veterans from top-tier silicon and automotive companies.Compensation to Match: can be cofounder-CTO position or first engineer (based for California: $320-350K and very competitive package).This job post was partly written by AI: complaints about mistakes please directly to LI/MS, Perplexity.