For many users, the most efficient way to access SpiNNaker is via the HBP Portal , through which you can submit jobs to the half-million core HBP SpiNNaker machine. The HBP system will schedule jobs, allocating as much machine resource as is required by each job.
To access SpiNNaker this way you will need to join the HBP Community and make a lightweight case for the use you require. Thereafter it is all free up to your allocated resource quota!
If you need a local SpiNNaker system, for example for real-time interaction with a physical robotics system, then the following options are available.
4-Node Board (72 processors)
The 4-node board is suitable for small-scale spiking neural networks, up to around 10,000 neurons (about the scale of a pond snail's brain!).
It requires a wired
Ethernet connection to a host machine (which can be a PC or Mac, including
laptops) and can be powered from a USB2 port on the host machine.
48-Node Board (864 processors)
The 48-node board is suitable for larger-scale spiking neural networks, up to a few 100,000s of neurons (equivalent to the scale of a small insect brain, such as drosophila).
It requires a wired Ethernet connection to a host machine (which can be a PC or Mac, including laptops) and has a mains adapter delivering 12V at up to around 5A.
The 48-node board has achieved CE approval.
24-Board Frame (20,736 processors)
The 24-board cage (shown here without its front screening cover) is suitable for spiking neural networks up to several million neurons (equavalent roughly to the scale of a small amphibian such as a frog).
It incorporates 24 48-node boards, power supply, cooling fans and a networks switch. It requires a mains electricity supply capable of delivering up to 2kW.
It requires a wired or optical Ethernet connection to a host machine, which ideally should have high computation power and memory to cope with the larger networks.
The 24-board cage can be configured with fewer boards - multiples of three make the most sense!
The 24-board cage has achieved CE approval with an optical host connection
and its front screening cover in place.
SpiNNaker Machine (518,400 processors)
The half-million core SpiNNaker machine (shown here with its cabinet doors removed) incorporates 25 24-board cages and is suitable for spiking neural networks up to hundreds of millions of neurons, equivalent to the scale of a few mouse brains.
Its hosting, cooling and power supply are our problem! It is configured as a single 2D toroidal surface and can therefore be used for a single large simulation, but we have software that can partition it into smaller units to support multiple smaller jobs that are physically isolated from each other (by disabling the inter-board communications at the periphery of each job). The host also runs each job in a separate virtual machine to minimise the risk of mutual interference.
This machine is openly available, at no cost, for on-line access to the HBP community.
We plan to expand this machine to a million cores over the coming months.
SpiNNaker Distribution Map
SpiNNaker has already achieved a world-wide user base. This map shows
countries where there are SpiNNaker machines - it does not show every
machine as that got too messy!
The total number of boards out there at
present is 90: 55 4-node and 35 48-node boards.
SpiNNaker is a novel computer architecture inspired by the working
of the human
brain whose development has been funded by the UK's Engineering and Physical
Sciences Research Council, EPSRC.
Where to go to find out more:
- Learn more about the SpiNNaker Project
- For more detail on the philosophy of the SpiNNaker Architecture
- The heart of the machine is the SpiNNaker chip
- The System Software running on the machine.
- To access tools and software to run on SpiNNaker systems, see our Downloads page
- Our Support page provides white papers, documents and FAQs.
- The Publications page gives details of papers describing SpiNNaker in detail
For further information on SpiNNaker development boards or the SpiNNaker project contact us at:
Our mail address is:
School of Computer Science,
University of Manchester
The design and construction of the SpiNNaker machine was funded by EPSRC
and The University of Manchester.
The ongoing support and software development, with provision of internet
access to the machine, is being supported by the EU through the ICT
Flagship Human Brain Project.
Research using the machine is being supported from the European Research
Council under the European Union's Seventh Framework Programme
(FP/2007-2013) ERC Grant Agreement no. 320689 BIMPC -
"Biologically-Inspired Massively-Parallel Computation".
The research has also received support from ARM Ltd, and from Samsung
through their GRO programme.
We are grateful to all these funding bodies and companies for their support.