Performance Engineer – Compute, Storage & AI Optimisation

Introducing evroc

At evroc, we are building a secure, sovereign, and sustainable hyperscale cloud to reimagine the digital future of Europe. By joining our company, you have an exciting opportunity to contribute to the development of next-generation cloud services. We are seeking intellectually curious and highly motivated team members who are ready to embrace the thrilling challenge of building the first European owned hyperscale cloud.

Location: Stockholm or London
Job type: Permanent position

Join the Pioneers

This is a rare opportunity to influence the technical foundation of a European hyperscaler— at the very beginning. As a Performance Engineer at evroc, you won’t just optimise infrastructure; you’ll shape it. Working with cutting-edge hardware, accelerators, and low-level software, you’ll ensure our compute and storage systems perform at their peak. From NUMA-aware tuning to GPU stack optimisation, your work will directly impact the performance of cloud-native services at scale. You’ll be part of a fast-moving infrastructure organization, collaborating closely with teams across hardware, kernel, and platform to deliver world-class cloud capabilities.

What You’ll Tackle

• Having a low-level understanding and a deep interest in Linux kernel optimisations and data paths, being able to understand and optimisie for performace

• Optimise NUMA configurations, CPU pinning, and kernel-level parameters across virtualized and bare metal systems

• Analyze and tune storage and I/O performance from physical devices through container and VM layers including network level NIC offloading, jumbo frame considerations, Virtualization pass-through for NICs and GPUs, object and block storage tuning, NVMe name spacing vs CPU/RAM utilization, block size optimisations dependent on disk characteristics, Kubernetes performance tuning, etcd optimisations etc

• Accelerate networking performance using technologies like SR-IOV, XDP, DPDK, and RDMA

• Tune NVIDIA GPU stacks including driver-level settings, CUDA runtime, NCCL, and MIG for hardware such as BG/GH200 as well as A100, H100 and LS40 and upcoming architectures

• Optimise interconnect topologies such as PCIe, NVLink, and NVSwitch, Infiniband, NVMe-oF, RoCE specifics etc

• Develop and maintain benchmarking, profiling, and performance tracking frameworks across compute, storage, AI infrastructure and networking

• Collaborate with cross-functional engineering teams to resolve system-level bottlenecks and inform roadmap decisions

Skills We Value

• Deep understanding of Linux systems internals, including CPU scheduling, memory management, storage I/O, and networking

• Experience with virtualization and hypervisors such as KVM/QEMU, libvirt, k8s, kubevirt, kubeovn and NUMA-aware VM placement

• Proficiency in performance analysis tools such as perf, eBPF, nvprof, Nsight, fio, and stress-ng

• Hands-on experience with NVIDIA GPU tooling and stacks (nvidia-smi, CUDA, NCCL, MIG)
  Familiarity with accelerator and interconnect technologies such as NVLink, PCIe, and InfiniBand

• Strong programming or scripting skills in languages such as Python, Go, or C
  Ability to thrive in fast-paced, collaborative infrastructure teams

Please note that you do not need experience with all technologies listed above to apply. If you have a strong background in systems performance and a drive to help build a better cloud, we’d love to hear from you.

We offer a competitive salary and an equity package to attract the best. Whether it’s the cosmopolitan allure of London, the vibrant Stockholm, or the Mediterranean charm of Sophia Antipolis, you’re poised for an inspiring work environment and a captivating local culture!

At evroc, diversity is our strength. We champion an inclusive environment where every background – ethnicity, age, gender identity, beliefs, and culture – is celebrated.

Applicants must possess a valid work permit.