Intel® Labs at Intel Innovation Digital

Highlights:

  • Intel Innovation digital takes place virtually from October 27-28th, 2021.

  • Intel Labs presents the latest developments in quantum computing, machine programming, AI, and neuromorphic computing.

  • AI Tech Insight Session: “The Fast Path to Scale AI Everywhere” to be presented by Intel Labs’ Parallel Computing Labs Director and Intel Senior Fellow, Pradeep Dubey.

BUILT IN - ARTICLE INTRO SECOND COMPONENT

Intel is proud to present its first annual Intel Innovation digital conference, a free virtual experience from October 27-28th, 2021. The conference features technical insights and training on today’s hottest topics, including AI, 5G, edge, cloud, and client solutions. 

Intel Labs will give in-depth presentations and demonstrations on several technology areas, including:

  •  Achieving quantum practicality
  • Using machine program systems to develop and maintain software
  • Responsible AI and human-AI collaboration
  • Scaling AI
  • Intel’s new neuromorphic technology, Loihi 2 and Lava, an open-source software framework.

Registration for Intel Innovation is now open at www.intel.com/innovation.

Intel Innovation is the first Intel ON event series event, which features ongoing events intended to inspire, engage, educate, and connect. The events offer the technology industry and academic community opportunities to learn about the latest thinking from business and technology innovators. 

To sign up to receive the latest updates and details about future Intel ON events, visit www.intel.com/oneventseries.

Following is a complete list of Intel Labs’ presentations and demonstrations at Intel Innovation: 

Wednesday, October 27th
12:00 PM - 12:30 PM PDT

Achieving Quantum Practicality with Silicon Spin-Qubits and a Full-Stack Quantum 
Anne Matsuura, Director, Quantum Applications & Architecture, Intel Corporation 

Intel’s quantum computing research aims to achieve quantum practicality where commercial quantum computing systems can tackle real-world problems that conventional computers simply can’t handle. However, there are numerous challenges to developing a practical quantum computing system. 

For example, qubits are extremely fragile, with any minute disturbance causing data loss. Moreover, there are significant issues of scale, with real-world implementations at commercial scale likely requiring at least one million qubits. Given that reality, Intel is actively developing design, modeling, packaging, and fabrication techniques to create silicon spin-qubits. 

This pioneering approach draws on Intel’s high-volume silicon manufacturing techniques to achieve greater reliability and scalability necessary for quantum practicality. In addition to hardware advances, Intel Labs has developed a full-stack quantum software development kit (SDK) for programmers to begin developing quantum algorithms. The SDK utilizes qubit simulators, including one that mimics Intel’s silicon spin-qubit hardware.

Wednesday, October 27th
12:00 PM - 12:30 PM PDT

A Giant Leap in Developer Productivity with Machine Programming 
Justin Gottschlich, Principal AI Scientist & Director/Founder of Machine Programming Research, Intel Corporation 

Software is integrated into almost everything we do. Yet, developing and maintaining software is time-consuming and error-prone. Intel’s goal with machine programming (MP) is to democratize the creation of software. While there’s much work to be done, you’ll see an overview of our growing body of MP systems that range from automatically detecting performance and correctness bugs to automatically determining the underlying semantics of the code.

Wednesday, October 27th
12:00 PM - 12:30 PM PDT

Amplifying Human Potential: Responsible AI and Human-AI Collaboration Research 
Lama Nachman, Director of Human & AI Systems Research Lab at Intel Corporation

While AI brings great potential to improve many aspects of our lives, it also raises concerns regarding employment, privacy, safety, and many others. It’s time to change the narrative from human-AI competition to human-AI collaboration. AI can amplify human potential to assist people in everyday life, from education to manufacturing to supporting our most vulnerable population. 

To enable AI to work closely with people, we need to develop contextually aware technologies built on robust and ethical AI. In this talk, we discuss Intel’s research in different applications utilizing multimodal sensing and probabilistic computing.

Thursday, October 28th
9:40 AM - 10:30 AM PDT

AI Tech Insight Session:
The Fast Path to Scale AI Everywhere 

Pradeep Dubey, Intel Senior Fellow, Intel Labs Director, Parallel Computing Lab

The impact of AI is rapidly growing. Yet, data practitioners continue to be held back by a common set of challenges. In this session, Pradeep Dubey and special guests share how Intel technology, software, and innovations are making AI more scalable, productive, and intelligent.

Thursday, October 28th 
11:00 AM PDT - 12:15 PM PDT

Entering a New Era of Neuromorphic Computing with Loihi 2 and Lava 
Mike Davies, Director, Neuromorphic Computing Lab, Intel Corporation
Andreas Wild, Research Scientist, Intel Corporation
Garrick Orchard, Research Scientist, Intel Labs
Yulia Sandamirskaya, Research Scientist, Intel Corporation

Neuromorphic computing promises to bring great advances to intelligent devices that must process real-time data streams. This type of computing must also adapt to unforeseen changes under power and latency constraints. Our first-generation chip called Loihi, announced four years ago, has now reached the hands of over 140 research groups globally. 

These research efforts have confirmed that neuromorphic technology can provide order-of-magnitude gains for many applications. These applications range from sensing and closed-loop control in edge devices to solving search and optimization problems in the data center. Loihi’s results have also exposed research challenges that impede progress to commercial viability. 

This lab introduces our second-generation chip, Loihi 2, and a new open-source software framework called Lava. Together, they build on the learnings from Loihi to hasten a new era of progress in neuromorphic computing. A series of demos using Lava gives you a hands-on view of the latest neuromorphic innovations.

Technology Showcase Demonstrations: 

  • Machine Programming 
    ControlFlag
    : a Self-Supervised Machine Learning Anomaly Detection System
     
  • Human-AI Collaboration
    Controllable Response Generation
    : Enabling real-time communication with minimal user intervention for persons with communication disabilities

Intel Innovation digital is sponsored by Accenture, Dell Technologies, Lenovo, Inspur, Microsoft, and VMware.