Software Performance Architect

Job Overview:

We are the CPU Technology team, part of Arm Central Technology Group. This diverse engineering-centric group invents, defines, and architects Technologies to be deployed in tomorrow's IP solutions from Arm.

The CPU Technology team tackles next-generation Arm CPUs, inventing new concepts that will fuel billions of devices worldwide!

Responsibilities:

• Analyse benchmarks and workloads to identify software and hardware optimization opportunities, analyse CPU performance and understand limitations we need to break through.
• Collaborate with various software teams, compiler, OS, applications to implement improvements.
• Collaborate with CPU design teams to identify bottlenecks.
• Write/Extract performance tests out of existing SW applications and run it on C/C++ models, emulators or real devices (Laptops, Smartphones, Servers…) on various operating systems.
• Work alongside modelling engineers, CPU design teams and Arm Architecture group.
• Drive technical activities.
• Collaborate with Business teams and partners to present technical results.

Required Skills and Experience:

• You graduated from a university or Engineering School, in Computer Science, Mathematics, Electronic / Electrical Engineering, or other related science.
• You have a bare minimum of several years of either software, hardware, or mathematic experience.
• You have some proven experience in SW programming alongside with technical management.
• You have a passion to innovate, think differently, and explore new avenues.
• You can work efficiently alone as well as in a team environment.

“Nice To Have” Skills and Experience:

• Having strong SW development experience in one or several domains like HPC, machine learning, distributed applications, web technologies, mobile applications, databases, multi-media.
• Having strong development skills in one or more high-level programming languages (C, C++, Java, Rust).
• Have SW performance analysis experience in platforms like Android (big/LITTLE systems), Windows, Linux, and operating system settings impact (power management, frequency governor), single-threaded vs multi-threaded.
• Having working experience in CPU performance analysis, methodology (PMU-based, TDA), tools (linux perf, Intel Vtune, android simpleperf/perfetto), aarch64 assembly language programming.
• Having knowledge on CPU architecture and micro-architecture performance techniques (branch prediction, prefetchers).
• You are used to develop and adapt some experimental application code to test new architectures.
• You demonstrate passion, drive, and diligence.
• You have good written and verbal communication skills.

In Return:

We work directly with engineers across the company to drive next-generation hardware. Your work will have a direct impact on our bottom line and the ability to deliver improvements for our customers. You will be part of a growing and fast-paced initiative within a team with varied strengths and give direction to your own work.