FuriosaAI

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.

AI AcceleratorsData CenterEfficiencyFull-Stack AILLM InferenceNPUSustainabilityTensor Contraction Processorfuriosa.ai

FuriosaAI

AI AcceleratorsData CenterEfficiencyFull-Stack AILLM InferenceNPUSustainabilityTensor Contraction Processorfuriosa.ai

HQUndisclosed

Team Size154

Open Jobs40

Total Funding$266M

Latest Fundraise8 months ago

Join the Team

Software Engineer, Containers and Kubernetes

HybridSeoul, Seoul, KR

Hybrid • Seoul, Seoul, KR

Teeming tracks opportunities at over 24,000 AI startups, then works with you to find (and land) the one you'll love.

DevOps Engineer

InternshipTel Aviv

Internship • Tel Aviv

Technical Writer

Full-timeRotterdam, NL

Full-time • Rotterdam, NL

Technical Writer

Full-timeSan Francisco, US

Full-time • San Francisco, US

DevOps Engineer

Full-timeTel Aviv

Full-time • Tel Aviv

Mobile Developer

Part-timeBelgrade, RS

Part-time • Belgrade, RS

Software Engineer

InternshipNovi Sad, RS

Internship • Novi Sad, RS

About The Job

FuriosaAI is looking for outstanding software engineers to design and build an end-to-end Kubernetes stack for next-generation AI chips.
In this role, you will work closely with the various teams for kubernetes integration, design, and build a resource scheduling and management solution at datacenter scale as part of a small team of experienced software engineers.
We are focused on supporting our products as first-class resource in kubernetes accross a variety of customer environments, both bare-metal and virtualized platforms. We expect you to have strong programming skills related to Go or Rust and deep understanding of kubernetes and linux container internals as well as excellent communication and planning skills.

Responsibilities

Integrate FuriosaAI NPU products as first-class resources in Kubernetes ecosystem.
Develop and maintain scheduling and resource management solution for datacenter scale.
Collaborate with teams working on ML applications, SDK, runtime, device driver, firmware and infrastructure.
Maintain high-quality software products through robust test coverage.

Minimum Qualifications

Bachelor’s degree in Computer Science/Engineering, or any related field, or a equivalent work experience
Strong understanding of Data Structures and Algorithms.
Proven ability to write high-quality, efficient, and maintainable code.

Preferred Qualifications

3+ years of strong experience with the Go or Rust programming language.
Solid understanding of Kubernetes Architecture, Container Runtime and Linux container internals.
Background in Cloud Native environment and distributed system design.
Experience of writing dynamic GPU Scheduler for MIG-Capable GPUs.
Knowledge of Hardware Topology Management and Optimization for NUMA or PCIE-Topology or NVLink/NVSwitch
Real world strong hands-on experience of Building and Operating GPU Farm or GPU Cluster.
Experience of writing Kubernetes Operator with kubebuilder, operator sdk or informer.

Contact

recruit@furiosa.ai