Sr. Principal Architect - AI Platform

Company Description

About Huawei

With 194,000 employees and operating in more than 170 countries and regions, Huawei is a leading global creator and provider of information and communications technology (ICT) infrastructure and smart devices. Integrated solutions span across four key domains – telecom networks, IT, smart devices, and cloud services. Huawei is committed to bringing digital to every person, home and organization for a fully connected, intelligent world.

About Huawei Canada

Huawei Canada focuses on fundamental research and development aimed at solving complex technical problems in emerging technologies like 5G, AI, Human Computer Interaction and Autonomous Driving. With ongoing research initiatives with 10 Universities across Canada and strategic collaboration agreements with several Universities, we support Canada’s rich research community.  In 2020, Huawei Canada ranked among the Top 20 corporate R&D investors in the country with a huge 40% increase in R&D investment year over year. Huawei Canada was established in 2008 and now has a total workforce of 1,200 in our six research centers across Canada.

Why work with Huawei Canada?

You will have the opportunity to work on real-world problems that impact people across the globe.  Many of our researchers are actively involved in publishing conference and journal papers, inventing patents and solving challenging technical problems. With cutting edge tools, access to highly specialized leaders and researchers, and significant funding, you will be well supported to fulfil your potential and pursue your professional dreams.

Job Description

We are looking for a talented, experienced, and motivated Senior Principal Architect to join our AI platform team. Our mission is to make the cutting-edge deep learning models achieve high accuracy and fast performance with Huawei's own AI strategy and full-stack portfolio, including a series of AI chips, cloud services, and products. As a Senior Principal Architect, you will also be leading the communication to both academic and industrial AI communities, to promote our research work and Huawei's AI Ecosystem. This AI Platform Senior Principal Architect role requires both broad and deep technology knowledge in this domain, will be responsible for defining the next generation of AI platforms to support Huawei Cloud & AI business.

Responsibilities

• Lead a team in research of algorithms on accelerating the training of the market-driven AI models on Huawei’s Ascend Platform, reaching/exceeding the state-of-the-art accuracy.
• Publish high-quality AI research papers, and attend conferences to promote Huawei’s AI Ecosystem;
• Track the trend of AI theory and technology development in the world and generate research report and proposals.
• Drive collaborations with other teams from Huawei’s global research centers.
• Plan project and technology roadmap.

Qualifications

• MSc, or PhD in Computer Science, Math/Statistics, focusing on AI & Deep Learning with solid publication records.
• 5+ years working experience in optimizing the performance of training deep learning models and/or their applications to CV/NLP/GNN domains.
• Expert knowledge of AI accelerators and the full stack of AI acceleration systems.
• Familiar with basic models (GPT series, switch-Transformer, etc.) and distributed training acceleration framework/technology (DeepSpeed，horovod，BytePS, etc.). 
• Strong math background in acceleration algorithms, such as gradient descent optimization, loss functions, new model architecture, mix precision, model compression, learning technologies (e.g., meta-learning), etc.
• Hands-on programming skills in C++/Python, experience with Tensorflow/PyTorch/Keras/MXNet.
• Excellent written and verbal communication skills for documentation and presentation.
• Team leader experience in AI industry is preferred.
• Be able to work with a global team, occasionally night meetings and travel across countries.  

Additional Information

We appreciate your interests and would like you to understand that only a short list of candidates will be contacted.