
FuriosaAI is an AI chip company developing data center accelerators focused on efficient AI inference for large language models (LLMs) and multimodal applications. Their flagship product, RNGD (Gen 2), is designed with a proprietary Tensor Contraction Processor (TCP) architecture to achieve high performance and efficiency, targeting lower total cost of ownership. The company also offers the Gen 1 Vision NPU for computer vision tasks. FuriosaAI provides a comprehensive software stack, including the Furiosa SDK, to optimize and deploy AI models. They emphasize programmability, efficiency, and ease of use, aiming to make AI computing more sustainable. The company has a global presence with R&D headquarters in Seoul, Korea, and hubs in the US and Germany.

FuriosaAI is an AI chip company developing data center accelerators focused on efficient AI inference for large language models (LLMs) and multimodal applications. Their flagship product, RNGD (Gen 2), is designed with a proprietary Tensor Contraction Processor (TCP) architecture to achieve high performance and efficiency, targeting lower total cost of ownership. The company also offers the Gen 1 Vision NPU for computer vision tasks. FuriosaAI provides a comprehensive software stack, including the Furiosa SDK, to optimize and deploy AI models. They emphasize programmability, efficiency, and ease of use, aiming to make AI computing more sustainable. The company has a global presence with R&D headquarters in Seoul, Korea, and hubs in the US and Germany.
About The Job Software Engineer (Inference Engine)는 FuriosaAI NPU에서 구동되는 대규모 언어모델 및 멀티모달 모델을 위한 고성능 추론 엔진을 개발하고 최적화합니다. 최신 추론 최적화 기술을 선도적으로 연구조사 하여 엔진에 적용하며, 컴파일러팀, 하드웨어팀과 긴밀한 협업을 통해 엔진의 성능을 고도화하는 역할을 수행합니다. 본 직무의 결과물은 FuriosaAI SDK의 핵심 구성요소로서 고객의 AI 서비스 성능과 안정성, 그리고 성공에 직접적인 영향을 미칩니다.
Responsibilities
Preferred Qualifications
Contact