About The Position

Majestic Labs is a fast-moving US-Israeli AI startup building custom, full-stack AI servers for large-scale inference and training. The company is seeking a Researcher/Architect to characterize AI workloads, develop simulations, and translate architectural ideas into quantitative, defensible performance models. The role sits at the intersection of computer architecture, workload characterization, system-level modeling, and AI or data-intensive workloads. The researcher will extend and maintain system-level performance models, evaluate architectural trade-offs, and compare systems with existing and emerging AI platforms. The role involves working closely with hardware architects, compiler and runtime teams, and external partners, and supporting technical discussions with customers and strategic partners. It also involves explaining performance modeling clearly to executives and investors.

Requirements

  • A bachelor's degree in computer science, electrical engineering, computer engineering, or a similar field and 8 years of experience in computer architecture, systems research, or performance engineering; or a Ph.D. in a similar field and 2 years of relevant experience.
  • Strong background in performance modeling or simulation, including analytical or event-driven approaches.
  • Experience with AI or graph workloads.
  • Understanding of memory hierarchies, including NUMA, caches, and coherence; interconnects such as PCIe, fabrics, and RDMA; and GPU-based systems.
  • Hands-on coding experience in Python, C/C++, or a similar language.
  • Ability to reason from first principles rather than relying only on benchmarks.

Nice To Haves

  • Background in computer architecture simulators.
  • Exposure to AI/ML runtimes, compilers, or hardware-aware model design.
  • Research background, such as an MS or Ph.D. in computer systems, computer science, electrical engineering, architecture, or a related field.

Responsibilities

  • Extend and maintain system-level performance models covering compute, memory, interconnect, and scheduling.
  • Model system behaviors such as NUMA effects, contention, coherence, and bandwidth saturation.
  • Characterize real workloads, including LLMs, GNNs, graph analytics, and tabular models.
  • Evaluate architectural trade-offs and produce quantitative comparisons with existing and emerging AI platforms.
  • Work closely with hardware architects, compiler and runtime teams, and external partners.
  • Ensure simulator assumptions align with hardware and software realities.
  • Contribute to technical deep dives with strategic partners and customers.
  • Translate performance modeling into clear narratives for executives and investors.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service