Intelligence Processing Unit
The use of machine learning to reimagine software applications and service development is exploding. Companies from every corner of the industry -- the biggest cloud service providers to corporate industrials to financial services to healthcare to retailers -- are exploring new ways of building products and services using data-centric learning models in place of traditional explicit programming. The drive to deliver more timely and more accurate results is compelling an ever greater need for specialized computing power. GPUs have been hailed as the solution to those computing needs but the industry is actively searching for a better, more efficient solution. Graphcore has that solution.
Graphcore’s Intelligence Processing Units, or IPUs, are specifically designed for artificial intelligence and compute-dense graph applications. These are not GPUs, Graphics Processing Units, but rather graph processors especially adept at the kinds of computations used in understanding relationships within a sea of data. Architecturally, IPUs looking nothing like GPUs. They offer performance, latency, and power efficiency advantages a GPU will be unable to match. Graphcore offers support for popular industry ML frameworks and a full tool suite for developers to innovate both within and outside those frameworks.
As an AI Applications Specialist, you will work to port and optimize machine learning and artificial intelligence applications using Graphcore’s Poplar™ software and IPU processors, enabling breakthroughs in this rapidly moving field.
You will work on AI and ML applications, create application notes and blog content, and work closely with Graphcore’s field teams, customers and partners to help them in understanding and getting the most from our Intelligence Processing Unit (IPU) technology. Having access to world leading compute resources, you will develop applications that push the boundaries of what is possible with machine learning today. You will also act as a senior technical figure within our product support organization, debugging customer issues and providing concise summaries and recommended fixes to our core engineering teams.
In your work you will support some of the world’s top machine learning innovators at deep learning research groups, at academic institutions, at innovative machine learning start-ups, at leading automotive companies, and at some of the world’s largest cloud and internet companies. You will need to develop a deep understanding of the IPU architecture and the associated Poplar™ tools and become familiar with leading machine learning frameworks.
You will need to develop a deep understanding of the state-of-the-art in artificial intelligence & machine learning domains and work with our customers to develop new techniques which exploit the unique features of our IPU architecture. We want you to become an industry thought leader on Graphcore technology and machine learning applications in the cloud, in automotive and in embedded applications. You should be interested and keen to present at industry conferences and will be able to back this up with written blogs and compelling content.
- Develop strong technical relationships with researchers and engineers at our customers and partners and help them to develop new algorithms and achieve breakthroughs in artificial intelligence
- Become a recognized expert on Graphcore’s IPU technology and Poplar™ tools and deliver compelling training to our customers and partners
- Field & resolve challenging/complex customer support issues
- Shepherd critical customer issues and provide timely advance warning of critical issues that need attention
- Become a thought leader on machine learning and advocate for Graphcore’s IPU technology
- Work with the Product Management and Engineering to ensure a good flow of customer feedback that can be incorporated into future products
- 4+ years of either machine learning development experience or related experience writing and optimizing applications in HPC, scientific libraries, compilers, digital signal processors or GPUs.
- Deep experience with C++ and in-depth knowledge of computer architectures, high performance programming and parallel programming
- Ability to multitask effectively in a fast-paced environment
- Action-oriented with strong analytical and problem-solving skills
- Keen interest to learn about the exciting new field of AI
- Comfortable in a customer-facing environment
- Strong written and oral communications skills with the ability to effectively interface with management and engineering
- Strong team-working and excellent interpersonal skills
- Masters or PhD in related computationally intensive science or engineering field
- Experience with C/C++, parallel programming and knowledge of computer architectures
- Experience working with modern deep learning software architecture and frameworks including: Tensorflow, MxNet, Caffe and/or PyTorch
- Excellent communication & presentations skills and comfortable in a customer-facing environment
- Experience working with accelerators such as GPUs, DSPs or FPGAs
- Experience in the AI/machine learning, cloud or automotive space
We welcome people of different backgrounds and experiences and are committed to building an inclusive work environment that makes Graphcore a great home for everyone. We are an equal opportunity employer and want to build a work environment where everyone is happy, productive and respectful so they can do their best work. If you have a disability or additional need that requires accommodation, just let us know.