F

FuriosaAI

Power-efficient AI inference accelerator chips

Algorithm – AI System Researcher

AI ResearchConfirmed live in the last 24 hours
Full-timeSeoul, South KoreaNo salary listedIn-person

About the Job

Requirements

  • A master's degree in AI/ML, Computer Architecture, or a related field, or equivalent experience.
  • Understand the operating principles and performance characteristics of LLM inference, including attention, KV cache, prefill/decode, batching, and multi-chip parallelism.
  • Be able to use Python and PyTorch to perform performance modeling, simulation, and data-driven analysis.
  • Be able to structure ambiguous problems independently and lead hypothesis-testing cycles.
  • Communicate analytical results and technical judgments clearly through documents and presentations.

Responsibilities

  • Identify major technical challenges in next-generation NPU serving, such as cluster-level serving systems and heterogeneous computing systems, and define research projects.
  • Explore the design space of serving architectures and algorithms using analytical modeling, roofline analysis, cost models, and internally developed simulators; quantitatively evaluate performance and cost to provide evidence for key decisions.
  • Develop serving algorithms, including scheduling, batching, KV cache management, and disaggregation; validate them through modeling, experiments, and implementation on NPU hardware; develop promising ideas into proofs of concept or prototypes and collaborate with the software organization toward productization.
  • Conduct the full research cycle from hypothesis formation through validation on physical hardware, with emphasis on modeling and analysis.
  • Explore design spaces broadly through modeling and simulation, and validate key hypotheses directly on the company's NPU.
  • Select tools as needed for each problem, which may involve simulator-based roofline analysis or implementation on NPU hardware.

Desired Qualifications

  • Experience analyzing or optimizing the performance of distributed systems, cluster-level software, or large-scale workloads.
  • Understand the internal architecture of LLM serving systems such as vLLM, SGLang, or TensorRT-LLM.
  • Experience with quantitative system evaluation such as roofline analysis, analytical performance modeling, or total cost of ownership analysis.
  • Experience optimizing or developing accelerator-based systems such as GPU or NPU systems.
  • Be familiar with using AI coding tools for development and research.

About the Company

FuriosaAI

Power-efficient AI inference accelerator chips

FuriosaAI develops AI accelerators for data centers. Its flagship RNGD chip is a power-efficient accelerator designed to boost performance per watt for running large language models and other demanding AI workloads. The hardware is configurable with multiple processing elements to fit different workloads, and it is paired with advanced compiler technology and profiling tools to identify bottlenecks and optimize efficiency. This combination of flexible hardware and software optimization differentiates FuriosaAI from competitors, focusing on maximizing performance while keeping energy use low. The company targets data centers and enterprises that need high-performance, energy-efficient AI solutions, and aims to make AI computing more sustainable and scalable by selling its chips and related software tools and pursuing long-term contracts with large-scale customers.

Company size
201-500
Company stage
Series D
Total funding
$887.5M
Headquarters
Seoul, South Korea
Founded
2017

Recently Posted Jobs

Explore more opportunities at FuriosaAI

See more jobs