FPGAworld 2020 postpone to 2021
On March 11, the WHO (World Health Organization) has declared the assessment of the risk of spread and impact of COVID-19 to very high at the global level and Characterized as ‘Global Pandemic’.
As your safety is our top priority and concern, the Organizing Committee (DTU, AFRY and FPGAworld) decided to postpone FPGAworld 2020 event to next year.
Welcome to FPGAworld Conference 2021 in
Stockholm September 14 and Copenhagen September 16
The FPGAworld Conference is an international forum for researchers, engineers, teachers, students, and hackers. It covers topics such as complex analog/digital/software FPGA SoC systems, FPGA/ASIC based products, educational & industrial cases and more. Registration for attendees is free and includes 2*coffee, lunch and go home drink.
2021: Program for Stockholm and Copenhagen
Preliminary Keynote Speaker Copenhagen 2021
Keynote speaker: Tero Rissa, Xilinx, Finland
Title: Adaptable Domain-Specific Architecture (DSA) for AI inference
Abstract: Every day new AI models are being invented to achieve higher accuracy or lower hardware requirements. However, fixed AI accelerators such as GPUs and ASICs struggle to meet real-life performance due to increasing complexity in network layers, layer-to-layer dataflow and lower mixed precisions. In this talk, we will show various Domain Specific Architectures (DSA) to optimize CNN, LSTM, and MLP in different vectors such as throughput, latency, and hardware utilization.
CV: Tero has more than 20 years of industrial experience of scaling new technologies from research and concept engineering into commercial products. He started working with artificial neural networks in 2009 and has since held senior positions in both engineering and management. Most recently he has been AI/ML HW Acceleration Research Group Leader at Nokia Bell Labs, Chief Architect of Machine Learning for Nokia Mobile Networks, the lead architect of OZO in Nokia Technologies and director of R&D at Microsoft. Tero holds MSc in CS/EE, Digital and Computer Systems from TUT Finland, and Ph.D. in Computing from Imperial College London UK. He is also an emeritus member of the Nokia CEO Technology Council and Bell Labs Distinguished Member of Technical Staff.
Preliminary Keynote Speakers Stockholm 2021
Keynote speaker: Magnus Peterson, Synective Labs AB, Sweden
Title: FPGAs for Deep Learning applications – Trends, solutions, and tools
Abstract: Deep Learning has over the last years become a game-changer in many fields, adding extraordinary capabilities to systems and products. With strong competition from other technologies, FPGAs are in many cases a leading candidate platform for deep learning implementation, combining high performance, low power, and high flexibility. Seeing new high-volume opportunities, like automotive vision applications, the FPGA vendors have put large efforts into developing streamlined design flows for easy integration of deep learning inference engines into an FPGA design.
This talk will look at the current trends and present different ways of implementing Deep Learning inference on FPGAs. It will also walk you through the different approaches taken by the different FPGA vendors and their different deep learning toolsets.
CV: Magnus has more than 35 years of experience from developing high performance embedded solutions, mainly within vision and image processing, where FPGAs play a key role. He has been with Synective Labs since 2003 and worked previously with vision-based industrial inspection systems at Innovativ Vision AB.
Preliminary Sponsor, Exhibitors and Presenters 2021
Copenhagen and Stockholm 2021