Page 28 - Microelectronics and Semiconductor Materials at TU Dresden

SpiNNcloud: Computing Technology Inspired by the Brain

SpiNNaker2 refers to a chip for biologically inspired AI, Deep-tech startup SpiNNcloud Systems, a spin-off of
where the Chair of Highly-Parallel VLSI Systems and TUD, receives funding the European Innovation Council
Neuro-Microelectronics is transforming a single transis- (EIC) amounting to 2.5 million euros for its ground-
tor into a 70,000-chip cloud. breaking project “SpiNNode: SpiNNaker2 on the edge.”
It will enable SpiNNaker2 to be extended to mobile
applications, such as human-machine interaction, and
to be tested in realistic, industrial environments.

One challenge is the high energy consumption, which
has inspired researchers to get to work on the most
energy-efficient computing hardware for large-scale
applications, as this will be key to significantly reducing
the carbon footprint of AI.

The SpiNNcloud represents the world’s largest system Left: Prof. Christian Mayr / Image: Stefan Schiefer
for real-time AI with 0.4 ExaOps machine learning Right: The SpiNNcloud supercomputer for brain simulation and massi-
performance, millisecond latency and world-record vely parallel real-time AI / Image: Christian Mayr
energy efficiency. Application areas include smart cities,
robotics, 6G, tactile internet, industry 4.0 and autono- CONTACT
mous driving.
Prof. Christian Mayr
Chair of Highly-Parallel VLSI
Systems And Neuro-Microelec-
tronics// TU Dresden

✉ christian.mayr@ tu-dresden.de
tu-dresden.de/ing/elektrotechnik/iee/hpsn/

28