Software

Learn more about the options available in our software teams.

Central Engineering - Central Processing Unit (CPU) Central Engineering - Graphics Processing Unit (GPU) Central Engineering- Machine Learning (ML)

Central Engineering

Central Processing Unit (CPU)

(Part 1)

Location: Bristol, Cambridge

Arm CPUs are the heart of digital electronic devices. The CPU group develops and delivers all Arm CPU IP, from the applications processor in your phone to the real-time processor in your car and the microcontroller in every electronic gadget around you. We are a multidisciplinary group with several projects running in parallel, each project requiring hardware and software engineers with a variety of skills to work closely together to produce groundbreaking CPU products:

HW – Design: Build early prototypes, design micro-architecture, develop RTL using HDL languages (SystemVerilog) and scripting, and optimize PPAF (power, performance, area, frequency).
HW – Verification: Design and develop environments and test benches to verify functionality (architectural and micro-architectural), responsible for taking the design to tape out quality. Might use one or more techniques (e.g. simulation/formal, pseudo-random/directed), using programming and scripting knowledge (e.g. SystemVerilog, C++, Python, assembly).
HW – Physical Implementation: Develop flows to turn IP into actual chips using scripting languages (e.g. Tcl/Python) and EDA tools (e.g. Cadence/Synopsys), to enable RTL optimization and PPA (power, performance, area) benchmarking.
SW – Performance exploration & verification: Design and develop C++ software models of new generation CPUs to explore the design space and identify novel opportunities to improve CPU performance through micro-architectural enhancements. Interact with stakeholders to analyze CPU bottlenecks and to discuss, propose, and explore ideas that improve the micro-architecture. Further, using industry standard benchmarks. ensure that the new CPU designs meet the expected performance requirements.
SW – SVOS (System Validation on Operating Systems): Low-level software and CPU platform development and debug. Join us if you want to work at the software/hardware interface, you want to know more about how CPUs and operating systems work, you would like to develop C/Python tools that test CPUs by generating random instruction sequences, or you are interested in the hardware platforms used for system-level validation of functionality and performance, whether on simulation, emulation or FPGA.

Central Engineering

Central Processing Unit (CPU)

(Part 2)

Location: Bristol, Cambridge

About You

Our work is very collaborative and deeply technical, so we are looking for people who:

Enjoy working collaboratively as part of a team.
Enjoy finding creative solutions to complex technical problems.

In addition, a graduate role in the CPU group will be best suited to people who:

Are interested in and want to learn more about how computers/CPUs work.
Want to develop skills in scripting, logic design, verification, and/or software programming.
Study electronic or computer engineering and similar subjects, either as part of a degree or as personal projects.

What could you be doing in this role?

As a graduate in the CPU group, you will have real responsibilities from day one and you will get support and guidance from experts that will help you succeed. Your work will contribute to current and future Arm CPU developments and power the next generation of digital devices. Through teamwork, training, and dedication to personal development, you can start your developmental journey to become a first-class Arm engineer.

Central Engineering

Graphics Processing Unit (GPU)

(Part 1)

Location: Cambridge

Arm GPU is an exciting division, solving the tough task of bringing triple-A games with triple-A features to mobile. Making our GPUs - massively parallel systems with up to 24 cores - fit into a small chip within a mobile phone, while being powerful enough to run AAA games and heavy compute workloads, and also efficient enough to deliver good battery life, is a unique challenge.

Unlike our many CPUs, our GPU architecture is not public. This enables us to make big changes with each new GPU variant, delivering the latest graphics and compute technologies (such as VR/AR, AI, and raytracing), and allowing us to continuously innovate with significant performance and efficiency improvements.

GPU is one of the largest software groups within Arm, with teams covering compilers, DevOps, user space, and kernel device drivers, firmware, models, performance analysis, diagnostic and visualization tools, Open CL drivers, and many more.

As GPU development is heavily collaborative, many of our projects involve working with multiple teams, and offer a unique ability to get a flavor of many different technologies. Our projects are chosen to be as exciting as possible while solving real challenges, and provide a fantastic opportunity to learn about many different aspects of software development, compiler technology, operating systems/computer architecture, graphics, and more.

Central Engineering

Graphics Processing Unit (GPU)

(Part 2)

Location: Cambridge

About You

All code we work with is either C/C++ or Python-based, so being strong in one of more of those languages – particularly in C – is required. Some basic knowledge of OS concepts (such as memory management, scheduling, threads) is ideal. Some knowledge or experience with hardware concepts (e.g. via embedded programming) would be useful. Some knowledge of computer graphics concepts would be nice, but is not necessary.

What could you be doing in this role?

Graduates are invited to join teams across the organization – working with engineers to deliver interesting and impactful projects. As a member of a team, you will get a hands-on view of how Arm develops GPUs, and experience what it's like to work in a fast-paced tech company.

Central Engineering

Machine Learning

(Part 1)

Location: Cambridge

The Arm Machine Learning (ML) group is part of Central Engineering (CE-ML) and includes teams working on a varied range of topics: from driver development for specialized ML hardware, to fostering the ML ecosystem, to open source contributions for Arm products such as the Arm NN and Arm Compute (ACL) libraries, to analyzing ML performance on Arm IP, and more. We also contribute to community projects, such as TensorFlow and TVM. Most of the software developed in CE-ML is released as open source.

About You

The ideal candidate has an interest in ML and is keen to learn more about the cutting edge of technology in hardware, as well as software optimization for ML and AI. As most of the teams do not directly develop ML models, ML knowledge is not a requirement. We would instead appreciate an interest or knowledge in lower-level languages where memory management is important (such as C/C++). Requirements will vary, depending on the team.

ML candidates can expect to learn about deploying ML workloads across a range of Arm technology including CPUs, GPUs, and accelerators. Our teams write a broad spectrum of software from low-level drivers to graph compilers, high-level ML frameworks, and application code.  

Central Engineering

Machine Learning

(Part 2)

Location: Cambridge

What could you be doing in this role?

From research to proof-of-concept development, to deployment on ARM IPs, joining this team, would be a phenomenal opportunity to contribute to the full life cycle of ML projects and understand how state-of-the-art ML is used to solve real-word problems. 

Working closely with field experts in a truly multidiscipline environment, you will have the chance to explore existing or build new ML techniques, while helping unpick the complex world of use cases that are applied on high-end mobile phones, TVs, and laptops. 

Recent student projects include:

GPT-2 running on Android with ArmNN TFlite delegate.
ML performance investigations on the latest Arm silicon.