Lead SoC Architect (CPU/GPU)

This confidential startup company is developing next-generation high-performance compute silicon to accelerate the future of AI, machine learning, and large-scale computing. Backed by experienced technology leaders and an ambitious product vision, the company is building innovative CPU, GPU, and advanced SoC solutions designed to deliver breakthrough performance, scalability, and efficiency. Joining this team offers the opportunity to shape foundational architecture, influence product strategy from the earliest stages, and make a lasting impact in a fast-paced, highly collaborative environment. Technical Background Deep expertise in the architecture and development of high-performance CPU, GPU, or heterogeneous SoCs from architectural definition through silicon validation and high-volume production. Drive system architecture for next-generation compute platforms supporting AI training, LLM inference, HPC, graphics, and accelerated computing workloads. Define and optimize CPU/GPU microarchitecture, memory hierarchy, cache subsystems, interconnect fabrics (AMBA CHI, NoC, PCIe, CXL, UCIe), and system-level performance. Lead architectural performance modeling, simulation, workload analysis, and Power/Performance/Area (PPA) optimization across the SoC. Drive technical decisions across CPU/GPU subsystem integration, memory architecture, AI accelerators, chiplet-based designs, and scalable compute platforms. Qualifications M.S. or Ph.D. in Electrical Engineering, Computer Engineering, Computer Science, or related discipline. 10+ years of experience in semiconductor architecture with significant experience in CPU, GPU, AI accelerator, or high-performance SoC development. Proven track record leading multiple complex SoCs from architecture through tapeout, silicon bring-up, and high-volume production. Deep understanding of CPU or GPU microarchitecture, cache coherency, memory systems, interconnects, performance modeling, and system optimization. Experience with modern compute architectures including ARM, x86, or RISC-V is highly desirable. Experience designing silicon for AI/ML accelerators, HPC, graphics, or datacenter compute platforms is strongly preferred. #J-18808-Ljbffr SBT