Jump to content

Cognitive computer

From Wikipedia, the free encyclopedia
This is an old revision of this page, as edited by CyberXRef (talk | contribs) at 04:38, 14 July 2019 (merging SpiNNaker since that's a more suitable place). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

A cognitive computer combines artificial intelligence and machine-learning algorithms, in an approach which attempts to reproduce the behaviour of the human brain.[1] It generally adopts a Neuromorphic engineering approach. An example of a cognitive computer implemented by using neural networks and deep learning is provided by the IBM company's Watson machine. A subsequent development by IBM is the TrueNorth microchip architecture, which is designed to be closer in structure to the human brain than the von Neumann architecture used in conventional computers.[1] In 2017 Intel announced its own version of a cognitive chip in "Loihi", which will be available to university and research labs in 2018.

Intel Loihi chip

Intel's self-learning neuromorphic chip, named Loihi, perhaps named after the Hawaiian seamount Loihi, offers substantial power efficiency designed after the human brain. Intel claims Loihi is about 1000 times more energy efficient than the general-purpose computing power needed to train the neural networks that rival Loihi's performance. In theory, this would support both machine learning training and inference on the same silicon independently of a cloud connection, and more efficient than using convolutional neural networks (CNNs) or deep learning neural networks. Intel points to a system for monitoring a person's heartbeat, taking readings after events such as exercise or eating, and uses the cognitive computing chip to normalize the data and work out the ‘normal’ heartbeat. It can then spot abnormalities, but also deal with any new events or conditions.

The first iteration of the Loihi chip was made using Intel's 14 nm fabrication process, and houses 128 clusters of 1,024 artificial neurons each for a total of 131,072 simulated neurons.[2] This offers around 130 million synapses, which is still a rather long way from the human brain's 800 trillion synapses, and behind IBM's TrueNorth, which has around 16 billion by using 64 by 4,096 cores.[3] Loihi is now available for research purposes among more than 40 academic research groups as a USB form factor.[4] [5]

IBM TrueNorth Chip

DARPA SyNAPSE board with 16 TrueNorth chips

TrueNorth was a neuromorphic CMOS integrated circuit produced by IBM in 2014.[6] It is a manycore processor network on a chip design, with 4096 cores, each one having 256 programmable simulated neurons for a total of just over a million neurons. In turn, each neuron has 256 programmable "synapses" that convey the signals between them. Hence, the total number of programmable synapses is just over 268 million (228). Its basic transistor count is 5.4 billion. Since memory, computation, and communication are handled in each of the 4096 neurosynaptic cores, TrueNorth circumvents the von-Neumann-architecture bottleneck and is very energy-efficient, with IBM claiming a power consumption of 70 milliwatts and a power density that is 1/10,000th of conventional microprocessors.[7] The SyNAPSE chip operates at lower temperatures and power because it only draws power necessary for computation.[8]

SpiNNaker

SpiNNaker (Spiking Neural Network Architecture) is a massively parallel, manycore supercomputer architecture designed by the Advanced Processor Technologies Research Group (APT) at the School of Computer Science, University of Manchester.[9] It is composed of 57,600 ARM9 processors (specifically ARM968), each with 18 cores and 128 MB of mobile DDR SDRAM, totalling 1,036,800 cores and over 7 TB of RAM.[10] The computing platform is based on spiking neural networks, useful in simulating the human brain (see Human Brain Project).[11][12][13][14][15][16][17][18][19]

The completed design is housed in 10 19-inch racks, with each rack holding over 100,000 cores.[20] The cards holding the chips are held in 5 Blade enclosures, and each core emulates 1000 Neurons.[20] In total, the goal is to simulate the behavior of aggregates of up to a billion neurons in real time.[21] This machine requires about 100 kW from a 240 V supply and an air-conditioned environment.[22]

SpiNNaker is being used as one component of the neuromorphic computing platform for the Human Brain Project.[23][24]

On October 14, 2018 the HBP announced that the million core milestone had been achieved.[25][26]

Criticism

There are many approaches and definitions for a cognitive computer,[27] and other approaches may be more fruitful than the others.[28]

Specifically, there are critics who argue that a room-sized computer - like the case of Watson - is not a viable alternative to a three-pound human brain.[29] Some also cite the difficulty for a single system to bring so many elements together such as the disparate sources of information as well as computing resources.[30] During the 2018 World Economic Forum, there are experts who claim that cognitive systems could adopt the biases of their developers and this was demonstrated in the case of the Google image-recognition or computer vision algorithm, which identified African Americans unfavorably.[31]

See also

References

  1. ^ a b Dharmendra Modha (interview), "A computer that thinks", New Scientist 8 November 2014, Pages 28-29
  2. ^ "Why Intel built a neuromorphic chip". September 29, 2017. www.ZDNet.com
  3. ^ "Intel unveils Loihi neuromorphic chip, chases IBM in artificial brains". October 17, 2017. AITrends.com
  4. ^ https://www.top500.org/news/intel-ramps-up-neuromorphic-computing-effort-with-new-research-partners/
  5. ^ http://niceworkshop.org/wp-content/uploads/2018/05/Mike-Davies-NICE-Loihi-Intro-Talk-2018.pdf
  6. ^ Merolla, P. A.; Arthur, J. V.; Alvarez-Icaza, R.; Cassidy, A. S.; Sawada, J.; Akopyan, F.; Jackson, B. L.; Imam, N.; Guo, C.; Nakamura, Y.; Brezzo, B.; Vo, I.; Esser, S. K.; Appuswamy, R.; Taba, B.; Amir, A.; Flickner, M. D.; Risk, W. P.; Manohar, R.; Modha, D. S. (2014). "A million spiking-neuron integrated circuit with a scalable communication network and interface". Science. 345 (6197): 668. doi:10.1126/science.1254642. PMID 25104385.
  7. ^ http://spectrum.ieee.org/computing/hardware/how-ibm-got-brainlike-efficiency-from-the-truenorth-chip How IBM Got Brainlike Efficiency From the TrueNorth Chip
  8. ^ "Cognitive computing: Neurosynaptic chips". IBM. 11 December 2015.
  9. ^ Advanced Processor Technologies Research Group
  10. ^ "SpiNNaker Project - The SpiNNaker Chip". apt.cs.manchester.ac.uk. Retrieved 2018-11-17.
  11. ^ SpiNNaker Home Page, University of Manchester, retrieved 11 June 2012
  12. ^ Furber, S. B.; Galluppi, F.; Temple, S.; Plana, L. A. (2014). "The SpiNNaker Project". Proceedings of the IEEE. 102 (5): 652–665. doi:10.1109/JPROC.2014.2304638.
  13. ^ Xin Jin; Furber, S. B.; Woods, J. V. (2008). "Efficient modelling of spiking neural networks on a scalable chip multiprocessor". 2008 IEEE International Joint Conference on Neural Networks (IEEE World Congress on Computational Intelligence). pp. 2812–2819. doi:10.1109/IJCNN.2008.4634194. ISBN 978-1-4244-1820-6.
  14. ^ A million ARM cores to host brain simulator News article on the project in the EE Times
  15. ^ Temple, S.; Furber, S. (2007). "Neural systems engineering". Journal of the Royal Society Interface. 4 (13): 193–206. doi:10.1098/rsif.2006.0177. PMC 2359843. PMID 17251143. A manifesto for the SpiNNaker project, surveying and reviewing the general level of understanding of brain function and approaches to building computer modelof the brain.
  16. ^ Plana, L. A.; Furber, S. B.; Temple, S.; Khan, M.; Shi, Y.; Wu, J.; Yang, S. (2007). "A GALS Infrastructure for a Massively Parallel Multiprocessor". IEEE Design & Test of Computers. 24 (5): 454. doi:10.1109/MDT.2007.149. A description of the Globally Asynchronous, Locally Synchronous (GALS) nature of SpiNNaker, with an overview of the asynchronous communications hardware designed to transmit neural 'spikes' between processors.
  17. ^ Navaridas, J.; Luján, M.; Miguel-Alonso, J.; Plana, L. A.; Furber, S. (2009). "Understanding the interconnection network of SpiNNaker". Proceedings of the 23rd international conference on Conference on Supercomputing - ICS '09. p. 286. CiteSeerX 10.1.1.634.9481. doi:10.1145/1542275.1542317. ISBN 9781605584980. Modelling and analysis of the SpiNNaker interconnect in a million-core machine, showing the suitability of the packet-switched network for large-scale spiking neural network simulation.
  18. ^ Rast, A.; Galluppi, F.; Davies, S.; Plana, L.; Patterson, C.; Sharp, T.; Lester, D.; Furber, S. (2011). "Concurrent heterogeneous neural model simulation on real-time neuromimetic hardware". Neural Networks. 24 (9): 961–978. doi:10.1016/j.neunet.2011.06.014. PMID 21778034. A demonstration of SpiNNaker's ability to simulate different neural models (simultaneously, if necessary) in contrast to other neuromorphic hardware.
  19. ^ Sharp, T.; Galluppi, F.; Rast, A.; Furber, S. (2012). "Power-efficient simulation of detailed cortical microcircuits on SpiNNaker". Journal of Neuroscience Methods. 210 (1): 110–118. doi:10.1016/j.jneumeth.2012.03.001. PMID 22465805. Four-chip, real-time simulation of a four-million-synapse cortical circuit, showing the extreme energy efficiency of the SpiNNaker architecture
  20. ^ a b Video interview by computerphile with Steve Furber
  21. ^ "SpiNNaker Project - Architectural Overview". apt.cs.manchester.ac.uk. Retrieved 2018-11-17.
  22. ^ "SpiNNaker Project - Boards and Machines". apt.cs.manchester.ac.uk. Retrieved 2018-11-17.
  23. ^ Calimera, A; Macii, E; Poncino, M (2013). "The Human Brain Project and neuromorphic computing". Functional Neurology. 28 (3): 191–6. PMC 3812737. PMID 24139655.
  24. ^ Monroe, D. (2014). "Neuromorphic computing gets ready for the (really) big time". Communications of the ACM. 57 (6): 13–15. doi:10.1145/2601069.
  25. ^ "SpiNNaker brain simulation project hits one million cores on a single machine". Retrieved 2018-10-19.
  26. ^ Petrut Bogdan (2018-10-14), SpiNNaker: 1 million core neuromorphic platform, retrieved 2018-10-19
  27. ^ Schank, Roger C.; Childers, Peter G. (1984). The cognitive computer: on language, learning, and artificial intelligence. Addison-Wesley Pub. Co. ISBN 9780201064438.
  28. ^ Wilson, Stephen (1988). "The Cognitive Computer: On Language, Learning, and Artificial Intelligence by Roger C. Schank, Peter Childers (review)". Leonardo. 21 (2): 210. ISSN 1530-9282. Retrieved 13 January 2017.
  29. ^ Neumeier, Marty (2012). Metaskills: Five Talents for the Robotic Age. Indianapolis, IN: New Riders. ISBN 9780133359329.
  30. ^ Hurwitz, Judith; Kaufman, Marcia; Bowles, Adrian (2015). Cognitive Computing and Big Data Analytics. Indianapolis, IN: John Wiley & Sons. p. 110. ISBN 9781118896624.
  31. ^ Choudhury, Saheli Roy (2018-09-18). "A.I. has a bias problem that needs to be fixed: World Economic Forum". CNBC. Retrieved 2018-10-12.

http://www.foxnews.com/tech/2018/01/09/ces-2018-intel-gives-glimpse-into-mind-blowing-future-computing.html

Retrieved from "https://en.wikipedia.org/w/index.php?title=Cognitive_computer&oldid=906176673"