Deep Learning Hardware Engineer
Intel
About the Role
Intel's Client AI Architecture team is seeking a Deep Learning Hardware Engineer to help shape future generations of Intel's Neural Processing Unit (NPU) architecture for AI PCs, Edge AI, Agentic AI, Physical AI, and multimodal foundation models. This role operates at the intersection of AI, computer architecture, and hardware-software co-design to optimize AI workloads and drive innovations in next-generation AI acceleration technologies. You will collaborate with architecture, software, compiler, and machine learning teams to develop high-performance, power-efficient AI solutions for Intel client platforms. The ideal candidate has experience developing, optimizing, or deploying AI workloads on NPUs, GPUs, AI accelerators, or related computing architectures. Experience with Intel NPUs, OpenVINO, AI software stacks, compiler technologies, AI inference optimization, and hardware-software co-design is highly desirable.
Key Responsibilities
- Contribute to the design and optimization of neural compute and AI accelerator architectures.
- Develop and evaluate RTL implementations of hardware features supporting machine learning workloads.
- Analyze AI models and workloads to identify performance, power, and efficiency opportunities.
- Partner with architecture, software, and algorithm teams to define future hardware capabilities.
- Support hardware-software co-design efforts to optimize AI solutions across the technology stack.
- Evaluate neural network mapping strategies for accelerated inference execution.
- Apply optimization techniques such as quantization, compression, pruning, and sparsity to improve AI efficiency.
- Utilize computer architecture principles including pipelining, parallelism, memory hierarchy, and data movement optimization.
- Participate in technical innovation activities, including invention disclosures, patents, conference papers, and publications.
Requirements & Tech Stack
- Digital Logic Design and RTL Design using System Verilog or Verilog.
- AI model architectures, layers, activation functions, and neural network topologies (e.g., Llama, Qwen, DeepSeek, Pi0.5).
- Experience working with Git workflows and repositories.
- Optimizing AI models for performance, such as quantization, pruning, and compression.
Qualifications
- Bachelor's degree in Computer Science, Electrical Engineering, or a STEM-related field with 3+ years of relevant experience; OR
- Master's degree in the same fields with 2+ years of experience; OR
- PhD degree in the same fields with 3+ months of educational or work experience.
Experience requirements are obtained through a combination of industry relevant job experience, internship experiences and/or schoolwork/classes/research.
Preferred Qualifications
- Familiarity with AI frameworks, neural computing operations, and hardware-accelerated inference techniques.
- Hands-on experience with modern deep learning topologies, including foundational models.
- In-depth knowledge of computer architecture concepts such as multi-core/multi-threading, memory hierarchy, and floating-point vs. fixed-point computations.
- Experience collaborating in cross-disciplinary teams to align hardware and software requirements.