303 jobs for “software engineer”
NVIDIA Cambridge
Serve as the primary technical contact for academic and research partners, driving adoption of NVIDIA's accelerated computing and AI stack. Work directly with labs and faculty on large language model workloads, integrating frameworks and libraries into research projects. Deliver hands-on training, workshops and lectures, and track emerging research trends to shape prototype solutions.
NVIDIA Reading
Acts as the primary technical contact for academic and research partners, embedding NVIDIA's frameworks and software stack into large-scale AI and HPC workloads. Drives joint research projects, delivers workshops and trainings, and feeds researcher needs back to engineering. Requires deep LLM lifecycle expertise across multi-node systems and involves around 30% travel.
Wayve London
Owns reliability of large-scale GPU, Kubernetes and distributed AI training infrastructure, turning recurring operational problems into durable engineering solutions. Improves observability, automation, deployment safety, capacity planning and incident learning, and builds GitOps delivery with Argo CD or Flux. Works across AI Platform, Compute & Data Platform and Cloud Infrastructure teams.
Wayve London
Owns the end-to-end architecture and technical roadmap for Remote Assistance, the real-time bridge between Wayve's AI Driver and human operators, spanning onboard vehicle integration, offboard services and operator HMI. Defines the vehicle-operator contract, fallback/MRM behaviour and latency budgets, and leads cross-org execution across autonomy, safety, platform and product teams to support L4 operation.
NVIDIA Italy, Remote
This role involves owning the fabric operations of NVIDIA's Cloud Partner (NCP) fleets, including NVLink/NVSwitch partition management, switch software and firmware lifecycle, and fabric validation. You will work closely with customers, partners, and cross-functional teams to ensure reliable and efficient network operations in large GPU estates.
NVIDIA Zürich, Switzerland
Research and develop techniques to GPU-accelerate deep learning, machine learning and other AI workloads, working with technical experts to analyse and optimise complex AI and HPC algorithms on modern CPU and GPU architectures. The role also involves publishing optimisation techniques via developer blogs or conferences and influencing next-generation hardware and software design.
NVIDIA United Kingdom
Acts as the primary technical contact for academic and research partners, embedding NVIDIA's frameworks and software stack into large-scale AI and HPC workloads. Drives joint research projects, delivers workshops and trainings, and feeds researcher needs back to engineering. Requires deep LLM lifecycle expertise across multi-node systems and involves around 30% travel.
NVIDIA Courbevoie, France
Engages directly with customers and partners to architect and validate large-scale GPU-accelerated HPC/AI platforms, spanning hardware handover through Day 2 production stability. Works tool-agnostically across upstream Kubernetes, KubeVirt, Slurm and observability stacks, owning cluster-wide validation, burn-in and fleet reliability targets.
NVIDIA United Kingdom
Engages directly with customers and partners to architect and validate large-scale GPU-accelerated HPC/AI clusters, spanning hardware handover through Day 2 production stability. Works tool-agnostically across upstream Kubernetes, KubeVirt, Slurm, Prometheus/Grafana and Cumulus/SONiC, owning burn-in testing, fleet reliability and enablement runbooks.
NVIDIA Germany
Engages directly with customers and partners to architect and validate large-scale GPU/HPC clusters, owning full-solution validation and Day 2 production stability at fleet scale. Works tool-agnostically across upstream Kubernetes, KubeVirt, Slurm, Prometheus/Grafana and Cumulus/SONiC, acting as hands-on technical leader and trusted advisor for assigned accounts.
Wayve London
Lead a team of ML engineers and applied scientists building synthetic data from generative world models like GAIA-3, turning them into a production engine for training-grade driving experience. Own the loop end to end, from checkpoint and large-scale GPU inference to artefacts landing in driving-model training, while setting technical direction on rig transfer, pose transfer and controllability.
