Performance Engineer - Open Source

Canonical is a leading provider of open source software and operating systems to the global enterprise and technology markets. Our platform, Ubuntu, is very widely used in breakthrough enterprise initiatives such as public cloud, data science, AI, engineering innovation and IoT. Our customers include the world's leading public cloud and silicon providers, and industry leaders in many sectors. The company is a pioneer of global distributed collaboration, with 1000+ colleagues in 70+ countries and very few office based roles. Teams meet two to four times yearly in person, in interesting locations around the world, to align on strategy and execution.

The company is founder led, profitable and growing.

We have created a new career path at Canonical for Performance Engineers who bring exceptional clarity to software performance, correctness and efficiency. We are adding these roles to every engineering team at the company.

As an open source platform, we have teams that work at every level of the software stack - from the kernel up through server and desktop applications, all the way to the web. We are adding roles with this focus in every engineering team in the company, to act as drivers of performance engineering skills, tools, culture and insights to our entire product range. We have also created a central performance engineering team, where we will drive shared tools, dashboarding capabilities, measurement frameworks, analytical capabilities and skills. We call this career focus 'percorr' because we have found that the same deep insights that enable sophisticated performance measurement and design also enable teams to improve their harnesses and frameworks for correctness. Understanding not only the application, but the runtime, the compiler, and the system, all the way down to the CPU, gives an engineer the ability to think creatively and with insight about the right things to measure, and the right ways to verify correctness. We have found that performance engineers will often improve the deepest and trickiest areas of our test harnesses, and unblock their teams to test and measure previously 'dark' aspects of their work.

We also have teams that work with every major silicon vendor and cloud, with a particular emphasis on performance optimisation of the entire stack for that architecture, CPU family or hypervisor. Our goal is to ensure that Ubuntu users can select whatever platform they want for their workloads and know they have access to the very best performance and efficiency across the entire stack from kernel to codec.

If they wish, we will enable performance and correctness engineers to move between products every two years, so that they have the opportunity to raise the bar for performance in an array of products over time, and also that teams have the benefit of fresh eyes on their processes, thinking, harnesses, dashboards and code. Our performance and correctness leads are expected to be truly exceptional individuals, and also leaders who are confident setting expectations of change, sharing insights and skills, running initiatives and programs. There is no way to move a meaningful codebase without also moving the people behind it.

This is a career path for perfectionists and precision artists. Whether you have a particular language focus - C, C++, Rust, Golang, Python, Java, Flutter-Dart or others - or a particular love of the kernel or hardware or CPU instruction set, whether you see yourself as driving perceived user experience through performance or optimising the efficiency of data centers to reduce carbon footprint, this process is the best way to find a role at Canonical that speaks to your precision, rigour, insight and drive.

Location: We have teams in every time zone, and can accommodate applications from almost any country.

Nice-to-have skills we value

• Low-level CPU insights and assembly language - understanding cache, registers, vectorization, memory
• Kernel performance insights - scheduling, context switching, hardware acceleration
• Distributed systems performance insights - tracing, debugging
• Compiler level performance insights - profile guided optimization, inlining, assembler
• Runtime performance insights - Golang, Python, JVM, NodeJS
• Database performance evaluation - plans, indexing, data modelling, tracing, monitoring
• Web browser performance insights - JS, web frameworks, tracing, CSS
• Performance measurement tooling - profiling, eBPF, D-Trace, flamegraphs
• Benchmarks and performance measurement tools - Phoronix Test Suite, others
• Test harnesses and capabilities in C, C++, Rust, Python, Golang, NodeJS, JavaScript
  

The role entails

• Drive software performance and correctness in a specific project
• Collaborate with colleagues and community around the globe
• Work on shared tools and infrastructure for performance measurement, analysis and reporting
• Teach, train and develop skills in your team for performance culture and design
• Strengthen test capabilities, harnesses and pipelines
• Analyze systems and code to understand performance and scalability
• Solve problems directly or through influence, leadership and skills development
• Review code and technical designs