Who We Are?
The Deep-Learning pre-enabling team is part of the global Software and Advanced Technologies Group.

We optimize DL topologies (neural networks) to the latest Intel CPUs and based on our performance analysis, we work with the HW architecture teams to improve the performance of future Intel's CPUs for deep learning.

Our SW work translates into a faster and more powerful CPU for four years into the future.

What Would You Do?

• In this job, you will design, implement and optimize deep learning software.
• You will maximize the performance of your software on current server CPUs and for future server CPUs, well before these are introduced to the public.
• Your SW will impact the design of the CPU HW, as the HW architects use it as a target to optimize the HW for.
• You will learn and master Deep-Learning performance at the undelaying framework libraries level.
• You will master SW performance analysis on silicon and simulation environment.

Qualifications

What we're looking for?

• Excellent BSc or MSc graduate at CS or CE
• at least 8 years of relevant experience
• Knowledge of scripting languages and C/C++.
• Background in multi-threading and low level performance programming.
• Motivation to learn new technologies.
• Assembly level programming.
• Deep learning background is not required, but is an advantage
• Experience in PyTorch or TensorFlow is an advantage

Inside this Business Group

Enable amazing computing experiences with Intel Software continues to shape the way people think about computing – across CPU, GPU, and FPGA architectures. Get your hands on new technology and collaborate with some of the smartest people in the business. Our developers and software engineers work in all software layers, across multiple operating systems and platforms to enable cutting-edge solutions. Ready to solve some of the most complex software challenges? Explore an impactful and innovative career in Software.

Work Model for this Role

This role will be eligible for our hybrid work model which allows employees to split their time between working on-site at their assigned Intel site and off-site.