Staff Engineer, Graphcore
2+ days a week in office
Advanced AI systems and services
Be an early applicant
Advanced AI systems and services
501-1000 employees
Be an early applicant
2+ days a week in office
501-1000 employees
Company mission
To make machine learning faster, easier and more intelligent.
Role
Who you are
- The ideal candidate brings deep experience with HPC and cloud-native infrastructure, thrives in fast-paced, loosely scoped environments, and leads with action, decisiveness, and teamwork
- Bachelor's degree or equivalent practical experience in a relevant subject
- Experience with RESTful API development
- Experience building, deploying, and operating containerized workloads using Kubernetes and container runtimes such as Docker or Podman
- Experience with managing production Kubernetes clusters and workloads
- Programming experience with Go
- Hands-on experience deploying and operating infrastructure using Infrastructure-as-Code, source code version control, and CI/CD automation tools (e.g. Terraform/OpenTofu, Ansible, GitLab, GitHub Actions, Git version control)
- Experience with Redfish for datacenter hardware management, telemetry, provisioning, and control
- Experience specifying, scoping, estimating and detailing work plans in an AGILE and SCRUM framework, including priorities, risks, issues, impacts and constraints
- Strong Linux systems engineering experience, including administration, automation, and scripting with Bash and Python
Desirable
- Experience with AI coding assistants (Codex, Claude, etc)
- Experience with Kubernetes operator development (Custom resources)
- Experience with High Performance Computing (HPC) environments using SLURM or similar batch workload solutions
- Experience with virtualized deployments and the technologies they rely on (e.g. Open vSwitch, KVM, QEMU)
- Experience with distributed object, block, and file storage (e.g., Ceph)
- Experience in end-to-end deployment automation and CI of containerized services. Complete automation of pipelines for build, test, deploy, manage, alert, destroy, rebuild
- Experience with solutions for monitoring and observability (e.g. Grafana, Prometheus, OpenSearch/ElasticSearch, Loki, Mimir, OpenTelemetry, Fluentd, Kafka)
- Experience with managed switch configuration (e.g. EOS, SONiC, DNOS)
- Experience with PyTorch for AI workloads
- Solid understanding of cloud and infrastructure technologies, including APIs, virtualization, networking, block storage, resource management, and monitoring systems
What the job involves
- We are looking for an experienced Staff Engineer to join our System Management team to help develop the critical interfaces used by internal and external customers to manage system state. Positioned between the hardware and customer workload, we are responsible for building the substrate upon which all other teams build upon
- As part of the System Management team, you’ll collaborate across teams to make complex systems work seamlessly and reliably
- While your primary focus will be on engineering excellence and system-level development, we value individuals who are versatile and hands-on — willing to contribute wherever needed, providing direct support and troubleshooting as necessary
- The System Management team, part of the Software Platform group, helps build Graphcore products into large-scale AI solutions for our customers
- The System Management Team is responsible for building the interfaces between hardware and AI software and frameworks as well as providing interfaces for public/private cloud use
- This takes the form of a rack management solution that abstracts complex hardware administrative duties. As one of the first teams to work with pre-release hardware and software it’s vital you are comfortable with unproven components and capable of problem solving solutions no matter what
- Ownership of software engineering efforts across the full SDLC, including implementation, automated testing, integration, and production readiness for the rack management solution
- Ownership of critical infrastructure with the need to drive issues to resolution while collaborating effectively across teams
- Configure and test new Graphcore AI hardware and systems using Continuous Deployment and Infrastructure-as-code in internal and external datacentres
- Work with our Datacenter Operations Engineers to maintain and operate the fleet of AI systems at peak performance
- Drive corrective actions for systems that are not operating correctly, working with DC operations and Graphcore Engineering as required
Our take
The hardware required to train AI models comes at a high cost, both financially and environmentally, producing millions of dollars in expenses and hundreds of tons of greenhouse gases. In response, Graphcore develops accelerator chips designed to expedite AI system training while significantly reducing power consumption.
AI accelerators like those crafted by Graphcore specialise in enhancing specific AI applications, such as artificial neural networks, deep learning, and machine learning. Utilising multicore designs and prioritising low-precision arithmetic or in-memory computing, these accelerators elevate the performance of large AI algorithms, achieving breakthrough results in natural language processing and computer vision.
Graphcore has ambitious plans to create an AI supercomputer, aiming to be one of the most powerful in its class. However, progress on this has extended beyond its planned 2024 debut, and the company was purchased by Softbank later that year. Though the acquisition by the Japanese multinational lowered Graphcore's valuation and highlighted structural problems in the local financing of UK tech pioneers, the buyout will undoubtedly aid the company in its ambitions to shape the future of AI computing.
Freddie
Company Specialist at Welcome to the Jungle
Insights
Company
Funding (last 2 of 6 rounds)
Dec 2020
$222m
SERIES E
Feb 2020
$150m
SERIES D
Total funding: $682m
Company benefits
- Work-Life Balance: We believe you should have a good work-life balance. Everyone's situation is different, so we offer flexible working arrangements and start and finish times, allowing you to choose when to work from home or come into the office
- Wellbeing: We want to ensure that our employees have the resources and support they need to prioritise their health and wellbeing
- Healthcare: We offer Private Medical Insurance (PMI) to our employees to help them take care of their health and that of their loved ones
- Progression: We prioritise learning and development as an integral part of our culture at Graphcore. Our Talent Development team is dedicated to equipping all our employees with the necessary tools, resources, and support to enhance their skills and reach their full potential in their careers
- Pension: We believe that our employees deserve the best when it comes to planning for their future
- Income Protection: Graphcore understands that unforeseen circumstances can happen, and we want to ensure our employees have the necessary financial security in place
Company HQ
Old City, Bristol, UK
Leadership
Nigel Toon
(Co-Founder & Exec Chair)Previously the President and CEO at picoChip Designs (processors for telecoms). As well as the VP and GM of Europe at Altera (semiconductors, merged with Intel).
Simon Knowles
(Co-Founder & CTO)Founded two successful semiconductor companies, Element14 and Icera, for a combined value of over $1 billion.
Salary benchmarks
We don't have enough data yet to provide salary benchmarks for this role.
Submit your salary to help other candidates with crowdsourced salary estimates.
Share this job
View 94 more jobs at Graphcore