HiPEAC 2022 embedded computing event returns to Budapest

The European embedded computing research community will be meeting again for the first time in-person at the HiPEAC 2022 conference in Budapest, Hungary, taking place from 20-22 June 2022. This will be the group’s first in-person event since HiPEAC in Bologna in January 2020.

HiPEAC 2020 Bologna
The last HiPEAC event took place in Bologna, Italy, in January 2020. (Image: Nitin Dahad)

HiPEAC is a European network of almost 2,000 world-class computing systems researchers, industry representatives and students. It forms a focal point for networking, dissemination, training, and collaboration activities in Europe for researchers, industry, and policy related to computing systems. HiPEAC’s mission is to advance computer architecture and computing systems research and development as a discipline in Europe.

The conference explores the low-level technology advances powering the latest developments in computing, from the most powerful high-performance machines to the tiniest devices at the edge. Three keynote talks, twenty-five workshops and five tutorials will complement the main paper track, along with sessions dedicated to industry and students. Its objectives include advancing computing systems for modern society, preparing the next generation of computer systems engineers, building a dynamic computing systems ecosystem for Europe, and aligning research efforts in computing systems and strengthen research impact in Europe – by identifying long-term challenges in computing systems and articulating their impact.

At HiPEAC 2022, each day will begin with a keynote from a technology expert.

Hai Li - Duke University
Hai Li (Image: Duke University)

On Monday 20 June, Hai Li (Duke University) will discuss efficient machine learning through algorithm-hardware co-design. She will look at machine learning and deep neural networks (DNNs); despite their widespread adoption in many applications, deployment of such big models, however, is both computation-intensive and memory-intensive. She argues that the progress of hardware development still falls far behind the upscaling of DNN models at software level. The holistic co-design across algorithm, architecture, and device levels emerges more important for execution acceleration, energy efficiency, and design flexibility. In her presentation, she presents their studies on how to optimize the training process for sparse and low-precision network models for general platforms. She also discusses the neural architectural search optimized for DNN operations.

ovidiu vermesan - sintef
Ovidiu Vermesan (Image: Sintef)

On Tuesday 21 June, Ovidiu Vermesan (SINTEF) will talk about edge distributed intelligence in the internet of autonomous systems. He outlines how the convergence of AI enabled edge processing, and the internet of things (IoT) enhances the cognitive capabilities of distributed autonomous systems offering new paradigms for addressing the collective behavior of self-organized autonomous systems. In this context, he argues that integrating AI-based methods such as ML and deep learning (DL) requires new concepts for energy-efficient and scalable solutions across the edge continuum. The presentation gives an overview of the future technology trends in the field, emphasizing the technology development roadmap for the following years.

Bianca Schroede_University of Toronto
Bianca Schroede (Image: University of Toronto)

On Wednesday 22 June, the keynote from Bianca Schroeder (Toronto University) will focus on flash-based storage in production systems. In this talk she looks at several years’ worth of field data to understand how these devices behave in production deployments, how they compare to HDDs and explore various aspects where reality might differ from expectations. In particular, she focuses on reliability and how it is affected by factors such as drive age, firmware, flash technology, and also other operational aspects, such as the level of write amplification experienced by SSDs in production systems and how it is affected by various factors; the effectiveness of wear leveling; or the rate at which drives in the field use up their program- erase (PE) cycle limit and what that means for the transition to future generations of flash with lower endurance.

The program of workshops complementing this main paper track, includes:

  • How to deliver high performance in cyber-physical systems at the edge, with real-time and safety constraints, for example the ‘safety-critical collaborative systems workshop trio’, the ‘mixed-critical systems workshop and tutorial’, the ‘next-generation real-time embedded systems workshop’ and the ‘de-RISC market-ready RISC-V-based platform for space applications’.
  • The latest developments in neuromorphic computing and architectures powering machine learning across the compute continuum, including the ‘BRAINE approach’, ‘deep learning for the IoT’ and ‘accelerators for machine learning’.
  • Inspirational examples of innovation across borders from the EU-funded BOWI and SMART4ALL initiatives.

Sponsors include Arm, DeepMind, Huawei and Cisco, the OpenHW Group, and sector-specific companies such as Cobham Gaisler, Collins Aerospace and Thales. Joining them in the industry exhibition and industry session are world-class European deep-tech companies, such as compiler experts Codeplay (recently acquired by Intel), dataflow engineer pioneers Maxeler (recently acquired by Groq), database maestros MonetDB and memory compression gurus ZeroPoint Technology. For the full list, see the 2022 HiPEAC conference webpages.

For students and companies attending the event in person, HiPEAC Jobs is organizing a range of activities under the umbrella of the STEM Student Day on 21 June. Drawing inspiration from the dynamic HiPEAC network, the STEM Day helps early career researchers find the most fulfilling careers, while assisting companies in finding the best candidates to deliver their future products.

Related Content:


Leave a Reply

Your email address will not be published. Required fields are marked *