Supercomputing, Software Engineer

Supercomputing, Software Engineer - Scheduling | OpenAICareersSupercomputing, Software Engineer - SchedulingSupercomputing - San FranciscoAbout the TeamThe Supercomputing Scheduling Pillar at OpenAI is dedicated to ensuring the reliability, scalability, and user-friendliness of job lifecycle management, with an emphasis on efficient and flexible job...

About The TeamSupercomputers scale vertically. The workloads are synchronous and cluster-scale. These conditions demand a novel approach to cluster infrastructure, and it is the work of the Supercomputing Scalability Pillar to invent it. The focus is on scaling beyond k8s supported node counts, deploying cluster wide releases rapidly and atomically,...

About the Team The Supercomputing Scheduling Pillar at OpenAI is dedicated to ensuring the reliability, scalability, and user-friendliness of job lifecycle management, with an emphasis on efficient and flexible job scheduling, quota management, and job execution workflows. We maximize researcher productivity by ensuring high goodput, efficient packing, and a...

Job DescriptionMagic’s mission is to build safe AGI that accelerates humanity’s progress on the world’s most important problems. We believe the most promising path to safe AGI lies in automating research and code to improve models and solve alignment more reliably than humans can alone. Our approach combines frontier-scale pre-training, domain-specific...

You will design, build, and maintain user interfaces that make managing large-scale job scheduling and cluster orchestration accessible and efficient. Your work will empower researchers and engineers to easily interact with and monitor complex supercomputing resources, focusing on usability and reliability.We’re looking for front-end engineers with...

About the Role You will design, build, and maintain user interfaces that make managing large-scale job scheduling and cluster orchestration accessible and efficient. Your work will empower researchers and engineers to easily interact with and monitor complex supercomputing resources, focusing on usability and reliability. We're looking for front-end...

Software Engineer - Power Management, Hardware HealthOpenAI’s Hardware Health team is dedicated to ensuring the optimal performance and reliability of our custom-built hyperscale supercomputers. We focus on maximizing supercomputing capacity for research and ensuring that our researchers are minimally impacted by hardware faults. This team is critical in...

About the Team OpenAI’s Hardware Health team is dedicated to ensuring the optimal performance and reliability of our custom-built hyperscale supercomputers. We focus on maximizing supercomputing capacity for research and ensuring that our researchers are minimally impacted by hardware faults. This team is critical in maintaining the infrastructure that...

About the TeamAt OpenAI, our Hardware Health team is dedicated to ensuring the optimal performance and reliability of our custom-built hyperscale supercomputers. We focus on maximizing supercomputing capacity for research and ensuring that our researchers are minimally impacted by hardware faults.The Hardware Health team operates within the broader Platform...

About the Team OpenAI's Hardware Health team is dedicated to ensuring the optimal performance and reliability of our custom-built hyperscale supercomputers. We focus on maximizing supercomputing capacity for research and ensuring that our researchers are minimally impacted by hardware faults. This team is critical in maintaining the infrastructure that...

The Platform Networking team is responsible for the collective communication stack used in our largest training jobs. Using a combination of C++ and CUDA we work on novel collective communication techniques that enable efficient training of our flagship models on our largest custom built supercomputers.The models we train are key ingredients to the AI...

About the Team The Platform Networking team is responsible for the collective communication stack used in our largest training jobs. Using a combination of C++ and CUDA we work on novel collective communication techniques that enable efficient training of our flagship models on our largest custom built supercomputers. The models we train are key ingredients...

About the TeamThe Platform Networking team is responsible for the collective communication stack used in our largest training jobs. Using a combination of C++ and CUDA, we work on novel collective communication techniques that enable efficient training of our flagship models on our largest custom-built supercomputers.About the RoleAs a Software Engineer,...

About the TeamThe Platform Runtime team builds the low-level framework components to power our ML training systems. We work on building robust, scalable, high-performance components to support our distributed training workloads. Our priorities are to maximize the productivity of our researchers and our hardware, with the goal of accelerating progress towards...

About the Team OpenAI's Hardware Health team is dedicated to ensuring the optimal performance and reliability of our custom-built hyperscale supercomputers. We focus on maximizing supercomputing capacity for research and ensuring that our researchers are minimally impacted by hardware faults. This team is critical in maintaining the infrastructure that...

AboutWe’re the San Francisco Compute Company. We’re building the first real-time compute trading platform. We think that over the next decade, thousands of startups and labs are going to be training and serving large models. They need compute to do this, and we’re building a platform on which that compute can be traded. If we’re successful, it will...

About the Team The Platform Runtime team builds the low level framework components to power our ML training systems. We work on building robust, scalable, high performance components to support our distributed training workloads. Our priorities are to maximize the productivity of our researchers and our hardware, with the goal of accelerating progress...

About the Team Storage Infrastructure provides APIs for data access, placement, and lifecycle management, while ensuring that the storage systems' capacity, throughput, and IOPs satisfy the needs of our AI researchers. Scalability, reliability, security, and usability are the core concerns of the team. About the Role As an engineer on the Storage...

About Eventual ComputingEventual Computing is a cutting-edge data platform that empowers data scientists and engineers to build scalable data applications across ETL, analytics, and ML/AI.We are on a mission to bridge the gap between traditional tabular data analytics and modern ML/AI workloads. Our open-source distributed data engine, Daft, runs on 800k CPU...

The Frontiers Infrastructure team builds the low level framework components to power our ML training systems. We work on building robust, debuggable, high performance libraries to support our distributed training workloads. Our priorities are to maximize the productivity of our researchers and our hardware, with the goal of accelerating progress towards...