Vice President, Architecture Research Lab
Samsung Semiconductor
San Jose, California, United States
1 day ago
Digital Design & SoC Architecture
AI/ML Hardware
$350,000 - $500,000 USD yearly
Originally posted November 20, 2026 by the employer.
About the role
This role leads research on memory-centric AI system architectures, covering the end-to-end co-design of AI workloads, system-level modeling, and hardware platforms. The objective is to design systems utilizing advanced memory solutions to address bottlenecks related to memory capacity, bandwidth, and interconnect.
What you'll do
- Define the research vision and strategy for memory-centric AI systems aligned with Samsung's memory technology roadmap.
- Manage multidisciplinary research teams of AI system architects and CPU architects, overseeing technical roadmaps for publications and patents.
- Establish system-level architectures to overcome memory capacity, bandwidth, and interconnect bottlenecks for large-scale AI workloads, including LLMs and recommendation systems.
- Direct rack-scale performance modeling, simulation, and design-space exploration across performance, power, and cost metrics.
- Oversee architecture path-finding and system specifications, including memory hierarchy design, for large-scale AI training and inference.
- Collaborate with SAIT HQ teams to integrate modeling insights into physical AI system implementations.
What you'll need
- Ph.D. in Computer Science, Electrical Engineering, or a related field.
- 10+ years of experience in system-level architecture research or large-scale computing platform design, with a focus on AI workloads.
- 10+ years of experience leading, building, and managing engineering teams.
- Proven track record of designing system-level architectures for large-scale AI inference from node up to rack or cluster scale, with deep knowledge of memory, networking, accelerators, and AI models.
- Comprehensive understanding of modern AI accelerator and system architectures (GPU, NPU, TPU) and heterogeneous integration strategies.
Nice to have
- Strong background in performance modeling and quantitative analysis of compute-memory-interconnect systems, including large-scale design-space exploration and PPA trade-off studies.
- Deep understanding of AI models (e.g., LLMs, MoE, recommendation models) with experience driving HW/SW co-design and HW-aware model optimization.
- Understanding of the inference serving and training software stack, with experience optimizing inference serving performance across multiple devices.
- Proven track record of publishing research papers in venues such as NeurIPS, ICML, MLSys, ISCA, MICRO, or HPCA.
- CPU ISA knowledge.
- Experience working in global, multi-geographic environments.
Skills: system architecture, memory-centric AI, AI accelerator, performance modeling, rack-scale
Market context
- Employer posting date: 20 Nov 2026
- First seen in our index: 20 Nov 2026 (our observation date)
- Last confirmed on the employer's career site: 20 Nov 2026
- Specialty: SoC/Chip Architecture (AI ML Hardware)
- Comparable active jobs: 35 at 12 employers, as of 20 Nov 2026
- Related specialties (Digital Design & Verification family): 3741 active jobs at 128 employers
View all Samsung Semiconductor jobs · SoC design jobs · San Jose jobs
How we counted: postings read from employer career sites; dates are our observations unless labelled as the employer's. How we count
Samsung Semiconductor
San Jose, California, United States
On-site
$350,000 - $500,000 USD yearly
1 day ago