Andrew Brown

Professor Andrew Brown (Southampton)
Electronics and Electrical Engineering, University of Southampton, SO17 1BJ, UK.
Room number: 3215 Building 5
email: adb@ecs.soton.ac.uk
Tel.: +44 (0)23 8059 3374
Fax: +44 2380 592901

I am a project collaborator with the School of Computer Science at the University of Manchester.

Biography

Andrew Brown currently holds an established chair in Electronics at the University of Southampton. He received a BSc in Physical Electronics from Southampton in 1976 and a PhD in Microelectronics in 1981. He held brief posts as research fellow and computer manager in the Electronics Department at Southampton before being appointed a lecturer at the end of 1980. He was promoted to Senior Lecturer in 1989, Reader in Electronics in 1992 and to one of the established chairs in 1999.
During his time as an academic, he has spent numerous secondments and sabbaticals working in industry. In 1983 he was appointed a Visiting Scientist in the Machine Technology group at IBM Hursley, UK, working on electronic place and route systems for uncommitted logic arrays. In 1985, along with three other academics, he founded Horus System Ltd, an EDA startup (backed by Cirrus Computers) to exploit simulation technology developed at the University. In 1988, he worked at Siemens NeuPerlach (Munich, Germany) on a micro-router for their in-house VENUS EDA suite. In 1995 he was awarded a Senior Academic in Industry secondment to work at a small communications company, MAC, developing a placement tool used in decision support for the placement of mobile phone base stations. In 2001, he co-founded LME Design Automation, a venture capital-backed spinout to exploit an EDA synthesis suite that was been the prime focus of his University research at that time. One consequence of this startup was that he was awarded a Royal Society Industrial Fellowship to continue his work there until 2003. In 2004, he was appointed a Visiting Professor at the University of Trondheim, Norway, and spent time there integrating the simulation and synthesis work of the previous two startup companies. In 2008 he was appointed a Visiting Professor at the Computing Laboratory, University of Cambridge, UK.
He was head of the Design Automation Research Group at Southampton from 1993 to 2007, when he became involved in the Manchester SpiNNaker system, and was able to obtain EPSRC support to allow him to work full time on the project, relinquish ing all teaching, supervision and management responsibilities.

Fellowships and professional qualifications

Andrew Brown is Fellow of the IET and BCS, a Senior Member of the IEEE, a Chartered Engineer and a registered European Engineer.

Technology exploitation

Horus Systems was set up in 1985 - backed by Cirrus Computers - to exploit the research carried out by the then Computer Group on simulation. LME Design Automation was set up in 2000 - backed by venture capital - to exploit the research carried out by the then EDA Group on synthesis. ECSpartners was set up in 2005 by the University to provide commercial consultancy services utilising the broad range of skills and capabilities available within the School of Electronics. ECS partners has worked with a broad range of organisations, from the smallest enterprise to many of the FTSE 100 and global companies. Professor Brown was one of the founding directors.

Professional service

Building a Common Vision for the UK Microelectronic Design Research Community. This community-building initiative began with a workshop hosted by the IEE at Savoy Place, London, in 2004. A series of further workshops and meetings led to a Network Grant proposal funded by EPSRC, which culminated in a set of Grand Challenge proposals. These were subsequently signposted by EPSRC.
eFutures - Maximising the Impact of UK electronics Research: The work of the Common Vision has been extended by the eFutures network grant - again supported by EPSRC - to encompass the devices community as well as microelectronics design and fabrication. Six participating Universities - Manchester, Southampton, Queens Belfast, Liverpool, Newcastle and Glasgow along with representatives from industry and EPSRC run the initiative (http://www.efutures.ac.uk).
UKDF The UK Design Forum is an informal annual meeting, usually hosted at the University of Manchester, with a mandate that encompasses all aspects of electronics design. This has been chaired since 2005 by Professor Brown.

General publications

A partial list of publications may be found at http://www.ecs.soton.ac.uk/people /adb/publications

Research interests

The trajectory of my research interests and activities has come full circle in the last thirty years. My PhD research focussed on electrochemical transducers[1], borne from an interest in gaining understanding of the nervous system by following the propagation of nerve impulses through neural tissue. The research was carried out under the aegis of the Microelectronics Group at Southampton, and required the design and fabrication of novel types of semiconductor device. This necessarily involved the production of integrated circuit masks, the technology of which was then (1976) in its infancy. Southampton was at that time one of the few Universities that possessed computers available at the research group level, so as an adjunct to the main research, I became involved in the design and development of what would now be called an Electronic Design Automation System. Naive by todays standards, this enabled us to replace the manual creation of masks (by cutting plastic sheets and photographically reducing them) with an automated system that allowed textual description of the mask geometry and automated mask production by writing directly into the photographic mask plates with a modified oscilloscope. This led to interest in the general problems of all aspects of electronic design automation: Work in mask verification[2] and network extraction[3] followed naturally from mask production (one being the inverse problem to the other), which led to place and route (the automatic translation of circuit structure to mask geometry), which in turn led to system synthesis[4] (the automatic translation of system behaviour to system structure).
Consistent with this has been work on electron beam proximity correction[5] (electron beams are used to write mask patterns on wafers during manufacture, but the electrons cause a kind of quantum-mechanical 'splash' when they land, blurring the image. Proximity correction is the problem of reverse engineering out the pattern you actually write, so that - post-splash - you end up with the pattern you wanted in the first place) and all aspects of simulation (analogue and discrete, the latter including neural networks[6]).
Numerous peripheral projects have arisen from this sequence of interests: power converter systems[7], magnetic systems[8] and transputer-based simulator implementation from simulation, ionospheric weather prediction from the electron beam work, and biometric otoacoustic emissions[9] and microfluidic sonoporation[10] from my long-standing relationship with the biosciences at Southampton. The simulation, place and route, multi core architectures and experience with building large software systems have come together over the last few years with my involvement with the SpiNNaker project: (http://apt.cs.man.ac.uk/projects/SpiNNaker/)
SpiNNaker - a precis: The human brain remains as one of the great frontiers of science - how does this organ upon which we all depend so critically, actually do its job? A great deal is known about the underlying technology - the neuron - and we can observe in-vivo brain activity on a number of scales through techniques such as magnetic resonance imaging, neural staining and invasive probing, but this knowledge - a tiny fraction of the information that is actually there - barely starts to tell us how the brain works, from a perspective that we can understand and manipulate. Something is happening at the intermediate levels of processing that we have yet to begin to understand, and the essence of the brain's information processing function probably lies in these intermediate levels. One way to get at these middle layers is to build models of very large systems of spiking neurons, with structures inspired by the increasingly detailed findings of neuroscience, in order to investigate the emergent behaviours, adaptability and fault-tolerance of those systems.
What has changed, and why could we not do this ten years ago? Multi-core processors are now established as the way forward on the desktop, and highly-parallel systems have been the norm for high-performance computing for a considerable time. In a surprisingly short space of time, industry has abandoned the exploitation of Moore's Law through ever more complex uniprocessors, and is embracing a 'new' Moore's Law: the number of processor cores on a chip will double roughly every 18 months. If projected over the next 25 years this leads inevitably to the landmark of a million-core processor system. Why wait?
We are building a system containing a million ARM9 cores - not dissimilar to the processor found in many mobile phones. Whilst this is not, in any sense, a powerful core, it possesses aspects that make it ideal for an assembly of the type we are undertaking. With a million cores, we estimate we can sensibly simulate - in real time - the behaviour of a billion neurons. Whilst this is less than 1% of a human brain, in the taxonomy of brain sizes it is certainly not a primitive system, and it should be capable of displaying interesting behaviour.
A number of design axioms of the architecture are radically different to those of conventional computer systems - some would say they are downright heretical. The architecture turns out to be elegantly suited to a surprising number of application arenas, but the flagship application is neural simulation; neurobiology inspired the design.
This biological inspiration draws us to two parallel, synergistic directions of enquiry; significant progress in either direction will represent a major scientific breakthrough:
"How can massively parallel computing resources accelerate our understan ding of brain function?"
How can our growing understanding of brain function point the way to mo re efficient parallel, fault-tolerant computation?

Technical challenges

SpiNNaker is not just another large computing system. It incorporates - at a fundamental level - a number of unorthodox design paradigms. It is designed primarily to simulate large aggregates of neurons (a billion). To do this, a million cores are interconnected by a novel communication infrastructure - details are in the publications and the SpiNNaker website. This involves distributing the topology of the network to be simulated throughout the topology of the processor network itself. The sheer size of both the simulating and simulated networks means that any central overseer - in almost any capacity - is not really feasible, and pretty much every aspect of the whole simulation ensemble has to be self-assembling. Factoring in the estimated mean time between failures intrinsic to extremely large systems compounds the technical challenges, because the simulating system has to be able to modulate its behaviour - on the fly - in the light of component and communication failures whilst a simulation is in progress.

• [1] A.D. Brown, 'An ionophoretic chemically sensitive transducer' Sensors a nd Actuators, 6, No 3, November 84, pp 151-168.
• [2] A.D. Brown and P.R. Thomas, 'Efficient Design Rule Checking Using a Sca nline Algorithm' Proceedings of the IEE I, 134, No 2, April 87, pp 63-69.
• [3] P.R. Thomas and A.D. Brown, 'A Goal-oriented subgraph isomorphism techn ique for IC device recognition'. Proceedings of the IEE-I, 135, No 6, December 88, pp 141-150.
• [4] A.C. Williams, A.D. Brown and M. Zwolinski 'Simultaneous optimisation of dynami c power, area and delay in behavioural synthesis', IEE Proceedings on Computers and Dig ital Techniques, 147, no 6, pp 383-390, Nov 2000.
• [5] C.S. Ea and A.D. Brown 'Enhanced pattern area density proximity effect correcti on', Journal of Vacuum Science Technology, pp 323-333, B17(2), March/April 1999, ISSN 0 734-211X.
• [6] E.T. Claverol, A.D. Brown and J.E. Chad, 'A large scale simulation of the pirif orm cortex by a cell automaton based network model', IEEE Transactions on Biomedical En gineering. 49, no 9, Sept 2002 pp 921-935
• [7] A.D. Brown, S.C. Wong, A.C. Williams and T.J. Kazmierski 'Fast time domain simu lation of a generic resonant mode power converter: mapping the stability region. IEE Pr oceedings on Circuits, Devices and Systems, 147, no 4, pp 211-218 Aug 2000.
• [8] A.D. Brown, J.N. Ross and K.G. Nichols 'Time domain simulation of mixed nonline ar magnetic and electronic systems', IEEE Transactions on Magnetics, 37, no 1, pp 522-5 32, Jan 2001.
• [9] Matthew A. Swabey, Paul Chambers, Mark E. Lutman, Neil M. White, John E. Chad, Andrew D. Brown and Stephen P. Beeby 'The biometric potential of transient otoacoustic emissions', Int. J. Biometrics, Vol. 1, no 3, 2009, pp 349-364
• [10] Rodamporn, S., Beeby, S., Harris, N., Brown, A., Hill, M. and Chad, J. 'Microfl uidic system for cell transfection using sonoporation and ultrasonic particle manipulat ion'. In: International Conference on Cellular & Molecular Bioengineering, 10-12 Decemb er 2007, Singapore.