77:
266:
605:
113:
36:
800:
541:
Hala Point packages 1,152 Loihi 2 processors produced on Intel 3 process node in a six-rack-unit chassis. The system supports up to 1.15 billion neurons and 128 billion synapses distributed over 140,544 neuromorphic processing cores, consuming 2,600 watts of power. It includes over 2,300 embedded x86
957:
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
649:
released Steve Lohr's article "What Ever
Happened to IBMâs Watson?". He wrote about some costly failures of IBM Watson. One of them, a cancer-related project called the Oncology Expert Advisor, was abandoned in 2016 as a costly failure. During the collaboration, Watson could not use patient data.
1035:
Song, Kyung Mee; Jeong, Jae-Seung; Pan, Biao; Zhang, Xichao; Xia, Jing; Cha, Sunkyung; Park, Tae-Eon; Kim, Kwangsu; Finizio, Simone; Raabe, Jörg; Chang, Joonyeon; Zhou, Yan; Zhao, Weisheng; Kang, Wang; Ju, Hyunsu; Woo, Seonghoon (March 2020). "Skyrmion-based artificial synapses for neuromorphic
1249:
Modha, Dharmendra S.; Akopyan, Filipp; Andreopoulos, Alexander; Appuswamy, Rathinakumar; Arthur, John V.; Cassidy, Andrew S.; Datta, Pallab; DeBole, Michael V.; Esser, Steven K.; Otero, Carlos Ortega; Sawada, Jun; Taba, Brian; Amir, Arnon; Bablani, Deepika; Carlson, Peter J. (2023-10-20).
452:. Intel claims Loihi is about 1000 times more energy efficient than general-purpose computing systems used to train neural networks. In theory, Loihi supports both machine learning training and inference on the same silicon independently of a cloud connection, and more efficiently than
532:
Intel's Loihi 2, named
Pohoiki Beach, was released in September 2021 with 64 cores. It boasts faster speeds, higher-bandwidth inter-chip communications for enhanced scalability, increased capacity per chip, a more compact size due to process scaling, and improved programmability.
463:. Intel points to a system for monitoring a person's heartbeat, taking readings after events such as exercise or eating, and using the chip to normalize the data and work out the ânormalâ heartbeat. It can then spot abnormalities and deal with new events or conditions.
1684:
Simon, George; DiNardo, Courtney D.; Takahashi, Koichi; Cascone, Tina; Powers, Cynthia; Stevens, Rick; Allen, Joshua; Antonoff, Mara B.; Gomez, Daniel; Keane, Pat; Suarez Saiz, Fernando; Nguyen, Quynh; Roarty, Emily; Pierce, Sherry; Zhang, Jianjun (June 2019).
1225:
551:
Hala Point integrates processing, memory and communication channels in a massively parallelized fabric, providing 16 PB S of memory bandwidth, 3.5 PB S of inter-core communication bandwidth, and 5 TB S of inter-chip bandwidth.
642:â is not a viable alternative to a three-pound human brain. 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.
1153:
Andreou, Andreas G.; Dykman, Andrew A.; Fischl, Kate D.; Garreau, Guillaume; Mendat, Daniel R.; Orchard, Garrick; Cassidy, Andrew S.; Merolla, Paul; Arthur, John; Alvarez-Icaza, Rodrigo; Jackson, Bryan L. (May 2016).
747:
Seo, Jae-sun; Brezzo, Bernard; Liu, Yong; Parker, Benjamin D.; Esser, Steven K.; Montoye, Robert K.; Rajendran, Bipin; Tierno, José A.; Chang, Leland; Modha, Dharmendra S.; Friedman, Daniel J. (September 2011).
253:
also announced its version of a cognitive chip in "Loihi, which it intended to be available to university and research labs in 2018. Intel (most notably with its
Pohoiki Beach and Springs systems),
545:
Intel claimed in 2024 that Hala Point was the worldâs largest neuromorphic system. It uses Loihi 2 chips. It is claimed to offer 10x more neuron capacity and up to 12x higher performance.
589:
548:
Hala Point provides up to 20 quadrillion operations per second, (20 petaops), with efficiency exceeding 15 trillion (8-bit) operations S W on conventional deep neural networks.
385:
In 2018, a cluster of TrueNorth network-linked to a master computer was used in stereo vision research that attempted to extract the depth of rapidly moving objects in a scene.
1226:"Inspired by the human brain â how IBM's latest AI chip could be 25 times more efficient than GPUs by being more integrated â but neither Nvidia nor AMD have to worry just yet"
358:, and computes by counting to a maximum of 19 bits. The cores are event-driven by using both synchronous and asynchronous logic, and are interconnected through an asynchronous
1850:
925:
401:. It blends approaches from IBM's 2014 TrueNorth system with modern hardware designs to achieve speeds about 4,000 times faster than TrueNorth. It can run
1431:
Tang, Guangzhi; Shah, Arpit; Michmizos, Konstantinos. (2019). "Spiking Neural
Network on Neuromorphic Hardware for Energy-Efficient Unidimensional SLAM".
445:
Pohoiki
Springs is a system that incorporates Intel's self-learning neuromorphic chip, named Loihi, introduced in 2017, perhaps named after the Hawaiian
566:
Loihi-based systems can perform inference and optimization using 100 times less energy at speeds as much as 50 times faster than CPU/GPU architectures.
1356:
1829:
1536:
428:
and can perform 2,048 operations per core per cycle at 8-bit precision, and 8,192 operations at 2-bit precision. It runs at between 25 and 425
1745:
1458:
1175:
769:
498:
555:
The system can process its 1.15 billion neurons 20 times faster than a human brain. Its neuron capacity is roughly equivalent to that of an
123:
49:
1383:
366:
142:
1762:
1635:
1610:
873:
178:
160:
63:
1081:
1314:
312:" that convey the signals between them. Hence, the total number of programmable synapses is just over 268 million (2). Its basic
1406:
1883:
222:
In 2023, IBM's proof-of-concept NorthPole chip (optimized for 2-, 4- and 8-bit precision) achieved remarkable performance in
1095:
900:
1559:
475:
1774:"The Cognitive Computer: On Language, Learning, and Artificial Intelligence by Roger C. Schank, Peter Childers (review)"
453:
1482:
Imam, Nabil; Cleland, Thomas A. (2020). "Rapid online learning and robust recall in a neuromorphic olfactory circuit".
494:
324:
Memory, computation, and communication are handled in each of the 4096 neurosynaptic cores, TrueNorth circumvents the
1008:
138:
1584:"Research Groups: APT - Advanced Processor Technologies (School of Computer Science - The University of Manchester)"
585:
466:
The first iteration of the chip was made using Intel's 14 nm fabrication process and houses 128 clusters of 1,024
365:
IBM developed a new network to program and use TrueNorth. It included a simulator, a new programming language, an
674:
413:
398:
325:
246:
208:
1652:
669:
402:
343:
The neurons are emulated using a Linear-Leak
Integrate-and-Fire (LLIF) model, a simplification of the leaky
196:
55:
1752:
1357:""Intel unveils Loihi neuromorphic chip, chases IBM in artificial brains". October 17, 2017. AITrends.com"
689:
679:
406:
370:
397:
for dramatically improving performance by intertwining compute with memory on-chip, thus eliminating the
1021:
750:"A 45nm CMOS neuromorphic chip with a scalable architecture for learning in networks of spiking neurons"
926:"Intel's Pohoiki Beach, a 64-Chip Neuromorphic System, Delivers Breakthrough Results in Research Tests"
449:
1360:
1338:
1263:
967:
721:
421:
355:
336:
chip operates at lower temperatures and power because it only draws power necessary for computation.
301:
943:
664:
1878:
1809:
1801:
1517:
1491:
1464:
1436:
1295:
1181:
1063:
1045:
991:
879:
828:
775:
578:
505:
490:
344:
294:
286:
204:
412:
tasks about 22 times faster, with 25 times less energy and 5 times less space, when compared to
377:
risks and other adverse consequences that may prevent it from commercialization in the future.
1793:
1758:
1724:
1706:
1660:
1631:
1606:
1509:
1454:
1287:
1279:
1171:
983:
869:
765:
639:
508:
published a research paper to demonstrate the ability of Intel's Loihi to recognize different
467:
409:
297:
265:
245:
microchip architecturewhich is designed to be closer in structure to the human brain than the
230:
223:
1583:
134:
1834:
1785:
1714:
1698:
1501:
1446:
1271:
1199:
1163:
1055:
975:
861:
757:
729:
482:. Loihi is available for research purposes among more than 40 academic research groups as a
394:
359:
313:
200:
1854:
560:
521:
509:
1267:
1125:
971:
725:
90:
Please help update this article to reflect recent events or newly available information.
1719:
1686:
659:
460:
425:
374:
308:
for a total of just over a million neurons. In turn, each neuron has 256 programmable "
234:
1251:
733:
604:
1872:
1521:
1299:
1067:
457:
417:
328:
bottleneck and is very energy-efficient, with IBM claiming a power consumption of 70
238:
27:
Machine designed to physically mimic the human brain and host artificial intelligence
1813:
1702:
1468:
1185:
995:
779:
1450:
1156:"Real-time sensory information processing using the TrueNorth Neurosynaptic System"
351:
280:
883:
855:
684:
1830:"Microchips that mimic the human brain could make AI far more energy efficient"
1687:"Applying Artificial Intelligence to Address the Knowledge Gaps in Cancer Care"
1433:
2019 IEEE/RSJ International
Conference on Intelligent Robots and Systems (IROS)
1167:
1155:
749:
207:
that closely reproduces the behavior of the human brain. It generally adopts a
1773:
1505:
1059:
761:
470:
each for a total of 131,072 simulated neurons. This offers around 130 million
329:
1797:
1710:
1664:
1283:
1275:
979:
574:
513:
479:
332:
and a power density that is 1/10,000th of conventional microprocessors. The
242:
1728:
1513:
1291:
987:
1754:
The cognitive computer: on language, learning, and artificial intelligence
1384:"Intel Ramps Up Neuromorphic Computing Effort with New Research Partners"
958:
integrated circuit with a scalable communication network and interface".
829:"IBM Debuts Brain-Inspired Chip For Speedy, Efficient AI - IEEE Spectrum"
582:
517:
446:
337:
254:
17:
1315:"NorthPole: Neural Inference at the Frontier of Energy, Space, and Time"
1805:
512:, which could eventually aid to "diagnose diseases, detect weapons and
471:
333:
309:
305:
272:
865:
588:
designed by the
Advanced Processor Technologies Research Group at the
1537:"Intel's Neuromorphic Loihi Processor Scales to 8M Neurons, 64 Cores"
1789:
1746:
CES 2018: Intel gives glimpse into mind-blowing future of computing
1496:
1441:
1050:
650:
Watson struggled to decipher doctorsâ notes and patient histories.
1200:"Stereo Vision Using Computing Architecture Inspired by the Brain"
801:"Samsung plugs IBM's brain-imitating chip into an advanced sensor"
429:
269:
264:
250:
1160:
2016 IEEE International
Symposium on Circuits and Systems (ISCAS)
373:
with any previous technology (e.g., C++ compilers) poses serious
833:
432:. This is an inferencing chip, but it cannot yet handle GPT-4.
283:
1082:"Neuromorphic computing: The long path from roots to real life"
857:
Smart
Machines: IBM's Watson and the Era of Cognitive Computing
636:
599:
556:
483:
290:
106:
70:
29:
1252:"Neural inference at the frontier of energy, space, and time"
635:
Critics argue that a room-sized computer â as in the case of
257:, and others are improving neuromorphic processors steadily.
712:
Witchalls, Clint (November 2014). "A computer that thinks".
1626:
Hurwitz, Judith; Kaufman, Marcia; Bowles, Adrian (2015).
1630:. Indianapolis, IN: John Wiley & Sons. p. 110.
1011:
How IBM Got Brainlike Efficiency From the TrueNorth Chip
590:
Department of Computer Science, University of Manchester
1560:"Intel Labs Launches Neuromorphic 'Kapoho Point' Board"
616:
340:
have been proposed as models of the synapse on a chip.
130:
754:
2011 IEEE Custom Integrated Circuits Conference (CICC)
241:
techniques. The following year, it developed the 2014
393:
In 2023, IBM released its NorthPole chip, which is a
141:, and by adding encyclopedic content written from a
420:that it was fabricated with. It includes 224 MB of
1096:"The brain's architecture, efficiency⊠on a chip"
901:"The brain's architecture, efficiency⊠on a chip"
823:
821:
493:published a research paper to demonstrate the
1751:Schank, Roger C.; Childers, Peter G. (1984).
304:, each one having 256 programmable simulated
8:
1603:Metaskills: Five Talents for the Robotic Age
1851:"New Chip Expands the Possibilities for AI"
1148:
1146:
707:
705:
577:(Spiking Neural Network Architecture) is a
64:Learn how and when to remove these messages
1628:Cognitive Computing and Big Data Analytics
1319:Dharmendra S. Modha - My Work and Thoughts
1120:
1118:
1116:
1022:"Cognitive computing: Neurosynaptic chips"
1718:
1495:
1440:
1049:
895:
893:
249:used in conventional computers. In 2017,
179:Learn how and when to remove this message
161:Learn how and when to remove this message
701:
542:processors for ancillary computations.
1849:Whitten, Allison (November 10, 2022).
474:, far less than the human brain's 800
122:contains content that is written like
1653:"What Ever Happened to IBM's Watson?"
1339:"Why Intel built a neuromorphic chip"
499:simultaneous localization and mapping
350:According to IBM, it does not have a
7:
854:KELLY, JOHN E.; HAMM, STEVE (2013).
1224:Afifi-Sabet, Keumars (2023-10-28).
1126:"IBM Research: Brain-inspired Chip"
1828:SERVICE, ROBERT F. (20 May 2022).
489:In October 2019, researchers from
367:integrated programming environment
25:
233:, a cognitive computer that uses
45:This article has multiple issues.
1605:. Indianapolis, IN: New Riders.
1313:Modha, Dharmendra (2023-10-19).
603:
111:
75:
34:
1703:10.1634/theoncologist.2018-0257
1382:Feldman, M. (7 December 2018).
53:or discuss these issues on the
1558:Peckham, Oliver (2022-09-28).
1451:10.1109/IROS40897.2019.8967864
1407:"Loihi - a brief introduction"
520:, and spot signs of smoke and
369:, and libraries. This lack of
1:
860:. Columbia University Press.
734:10.1016/S0262-4079(14)62145-X
454:convolutional neural networks
275:board with 16 TrueNorth chips
195:is a computer that hardwires
497:of Intel's Loihi in solving
362:mesh network on chip (NOC).
1535:Hruska, J. (16 July 2019).
1484:Nature Machine Intelligence
478:synapses, and behind IBM's
211:approach. Synonyms include
1900:
1757:. Addison-Wesley Pub. Co.
1651:Lohr, Steve (2021-07-16).
1168:10.1109/ISCAS.2016.7539214
944:"Korean Researchers Devel"
586:supercomputer architecture
1506:10.1038/s42256-020-0159-4
1060:10.1038/s41928-020-0385-0
762:10.1109/CICC.2011.6055293
559:brain or the cortex of a
504:In March 2020, Intel and
84:This article needs to be
1772:Wilson, Stephen (1988).
1601:Neumeier, Marty (2012).
1024:. IBM. 11 December 2015.
675:Neuromorphic engineering
326:von Neumann-architecture
247:von Neumann architecture
209:neuromorphic engineering
1588:apt.cs.manchester.ac.uk
1276:10.1126/science.adh1174
980:10.1126/science.1254642
670:Computational cognition
229:In 2013, IBM developed
197:artificial intelligence
1435:. pp. 4176â4181.
690:Spiking neural network
680:Tensor Processing Unit
399:Von Neumann bottleneck
371:backward compatibility
276:
1884:Unsupervised learning
268:
143:neutral point of view
1130:www.research.ibm.com
1412:. Intel Corporation
1405:Davies, M. (2018).
1268:2023Sci...382..329M
1084:. 15 December 2020.
972:2014Sci...345..668M
726:2014NewSc.224...28W
665:Cognitive computing
510:hazardous materials
416:which use the same
203:algorithms into an
135:promotional content
1657:The New York Times
1363:on August 11, 2021
1038:Nature Electronics
647:The New York Times
615:. You can help by
579:massively parallel
506:Cornell University
491:Rutgers University
468:artificial neurons
418:12-nm node process
389:IBM NorthPole chip
345:integrate-and-fire
300:design, with 4096
295:manycore processor
287:integrated circuit
277:
261:IBM TrueNorth chip
205:integrated circuit
193:cognitive computer
137:and inappropriate
1460:978-1-7281-4004-9
1262:(6668): 329â335.
1204:IBM Research Blog
1177:978-1-4799-5341-7
1132:. 9 February 2021
1100:IBM Research Blog
905:IBM Research Blog
884:10.7312/kell16856
866:10.7312/kell16856
771:978-1-4577-0222-8
633:
632:
495:energy efficiency
410:image recognition
298:network on a chip
293:in 2014. It is a
224:image recognition
213:neuromorphic chip
189:
188:
181:
171:
170:
163:
105:
104:
68:
16:(Redirected from
1891:
1865:
1863:
1861:
1845:
1843:
1842:
1835:Science magazine
1824:
1822:
1820:
1768:
1733:
1732:
1722:
1681:
1675:
1674:
1672:
1671:
1648:
1642:
1641:
1623:
1617:
1616:
1598:
1592:
1591:
1580:
1574:
1573:
1571:
1570:
1555:
1549:
1548:
1546:
1544:
1532:
1526:
1525:
1499:
1479:
1473:
1472:
1444:
1428:
1422:
1421:
1419:
1417:
1411:
1402:
1396:
1395:
1393:
1391:
1379:
1373:
1372:
1370:
1368:
1359:. Archived from
1353:
1347:
1346:
1335:
1329:
1328:
1326:
1325:
1310:
1304:
1303:
1246:
1240:
1239:
1237:
1236:
1221:
1215:
1214:
1212:
1211:
1196:
1190:
1189:
1162:. p. 2911.
1150:
1141:
1140:
1138:
1137:
1122:
1111:
1110:
1108:
1107:
1092:
1086:
1085:
1078:
1072:
1071:
1053:
1032:
1026:
1025:
1018:
1012:
1006:
1000:
999:
966:(6197): 668â73.
954:
948:
947:
946:. 30 March 2020.
940:
934:
933:
922:
916:
915:
913:
912:
897:
888:
887:
851:
845:
844:
842:
841:
825:
816:
815:
813:
811:
797:
791:
790:
788:
786:
756:. pp. 1â4.
744:
738:
737:
709:
628:
625:
607:
600:
436:Intel Loihi chip
395:proof-of-concept
316:is 5.4 billion.
314:transistor count
279:TrueNorth was a
201:machine learning
184:
177:
166:
159:
155:
152:
146:
124:an advertisement
115:
114:
107:
100:
97:
91:
79:
78:
71:
60:
38:
37:
30:
21:
1899:
1898:
1894:
1893:
1892:
1890:
1889:
1888:
1869:
1868:
1859:
1857:
1855:Quanta Magazine
1848:
1840:
1838:
1827:
1818:
1816:
1790:10.2307/1578563
1771:
1765:
1750:
1742:
1740:Further reading
1737:
1736:
1683:
1682:
1678:
1669:
1667:
1650:
1649:
1645:
1638:
1625:
1624:
1620:
1613:
1600:
1599:
1595:
1582:
1581:
1577:
1568:
1566:
1557:
1556:
1552:
1542:
1540:
1534:
1533:
1529:
1481:
1480:
1476:
1461:
1430:
1429:
1425:
1415:
1413:
1409:
1404:
1403:
1399:
1389:
1387:
1381:
1380:
1376:
1366:
1364:
1355:
1354:
1350:
1337:
1336:
1332:
1323:
1321:
1312:
1311:
1307:
1248:
1247:
1243:
1234:
1232:
1223:
1222:
1218:
1209:
1207:
1198:
1197:
1193:
1178:
1152:
1151:
1144:
1135:
1133:
1124:
1123:
1114:
1105:
1103:
1094:
1093:
1089:
1080:
1079:
1075:
1034:
1033:
1029:
1020:
1019:
1015:
1007:
1003:
956:
955:
951:
942:
941:
937:
924:
923:
919:
910:
908:
899:
898:
891:
876:
853:
852:
848:
839:
837:
827:
826:
819:
809:
807:
799:
798:
794:
784:
782:
772:
746:
745:
741:
720:(2994): 28â29.
711:
710:
703:
698:
656:
629:
623:
620:
613:needs expansion
598:
572:
561:capuchin monkey
539:
530:
522:carbon monoxide
461:neural networks
443:
441:Pohoiki Springs
438:
426:processor cores
391:
383:
360:packet-switched
322:
263:
235:neural networks
185:
174:
173:
172:
167:
156:
150:
147:
128:
116:
112:
101:
95:
92:
89:
80:
76:
39:
35:
28:
23:
22:
15:
12:
11:
5:
1897:
1895:
1887:
1886:
1881:
1871:
1870:
1867:
1866:
1846:
1825:
1769:
1763:
1748:
1741:
1738:
1735:
1734:
1697:(6): 772â782.
1691:The Oncologist
1676:
1643:
1636:
1618:
1611:
1593:
1575:
1550:
1527:
1490:(3): 181â191.
1474:
1459:
1423:
1397:
1374:
1348:
1330:
1305:
1241:
1216:
1191:
1176:
1142:
1112:
1087:
1073:
1044:(3): 148â155.
1027:
1013:
1001:
949:
935:
930:Intel Newsroom
917:
889:
874:
846:
817:
792:
770:
739:
700:
699:
697:
694:
693:
692:
687:
682:
677:
672:
667:
662:
660:AI accelerator
655:
652:
631:
630:
610:
608:
597:
594:
571:
568:
538:
535:
529:
526:
486:form factor.
442:
439:
437:
434:
390:
387:
382:
379:
375:vendor lock-in
354:, operates on
321:
318:
262:
259:
217:cognitive chip
187:
186:
169:
168:
139:external links
119:
117:
110:
103:
102:
83:
81:
74:
69:
43:
42:
40:
33:
26:
24:
14:
13:
10:
9:
6:
4:
3:
2:
1896:
1885:
1882:
1880:
1877:
1876:
1874:
1856:
1852:
1847:
1837:
1836:
1831:
1826:
1815:
1811:
1807:
1803:
1799:
1795:
1791:
1787:
1783:
1779:
1775:
1770:
1766:
1764:9780201064438
1760:
1756:
1755:
1749:
1747:
1744:
1743:
1739:
1730:
1726:
1721:
1716:
1712:
1708:
1704:
1700:
1696:
1692:
1688:
1680:
1677:
1666:
1662:
1658:
1654:
1647:
1644:
1639:
1637:9781118896624
1633:
1629:
1622:
1619:
1614:
1612:9780133359329
1608:
1604:
1597:
1594:
1589:
1585:
1579:
1576:
1565:
1561:
1554:
1551:
1538:
1531:
1528:
1523:
1519:
1515:
1511:
1507:
1503:
1498:
1493:
1489:
1485:
1478:
1475:
1470:
1466:
1462:
1456:
1452:
1448:
1443:
1438:
1434:
1427:
1424:
1408:
1401:
1398:
1385:
1378:
1375:
1362:
1358:
1352:
1349:
1344:
1340:
1334:
1331:
1320:
1316:
1309:
1306:
1301:
1297:
1293:
1289:
1285:
1281:
1277:
1273:
1269:
1265:
1261:
1257:
1253:
1245:
1242:
1231:
1227:
1220:
1217:
1205:
1201:
1195:
1192:
1187:
1183:
1179:
1173:
1169:
1165:
1161:
1157:
1149:
1147:
1143:
1131:
1127:
1121:
1119:
1117:
1113:
1101:
1097:
1091:
1088:
1083:
1077:
1074:
1069:
1065:
1061:
1057:
1052:
1047:
1043:
1039:
1031:
1028:
1023:
1017:
1014:
1010:
1005:
1002:
997:
993:
989:
985:
981:
977:
973:
969:
965:
961:
953:
950:
945:
939:
936:
931:
927:
921:
918:
906:
902:
896:
894:
890:
885:
881:
877:
875:9780231537278
871:
867:
863:
859:
858:
850:
847:
836:
835:
830:
824:
822:
818:
806:
802:
796:
793:
781:
777:
773:
767:
763:
759:
755:
751:
743:
740:
735:
731:
727:
723:
719:
715:
714:New Scientist
708:
706:
702:
695:
691:
688:
686:
683:
681:
678:
676:
673:
671:
668:
666:
663:
661:
658:
657:
653:
651:
648:
643:
641:
638:
627:
618:
614:
611:This section
609:
606:
602:
601:
595:
593:
591:
587:
584:
580:
576:
569:
567:
564:
562:
558:
553:
549:
546:
543:
536:
534:
528:Pohoiki Beach
527:
525:
523:
519:
515:
511:
507:
502:
500:
496:
492:
487:
485:
481:
477:
473:
469:
464:
462:
459:
458:deep learning
455:
451:
448:
440:
435:
433:
431:
427:
423:
419:
415:
411:
408:
404:
400:
396:
388:
386:
380:
378:
376:
372:
368:
363:
361:
357:
356:unary numbers
353:
348:
346:
341:
339:
335:
331:
327:
319:
317:
315:
311:
307:
303:
299:
296:
292:
288:
285:
282:
274:
271:
267:
260:
258:
256:
252:
248:
244:
240:
239:deep learning
236:
232:
227:
225:
220:
218:
214:
210:
206:
202:
198:
194:
183:
180:
165:
162:
154:
144:
140:
136:
132:
126:
125:
120:This article
118:
109:
108:
99:
87:
82:
73:
72:
67:
65:
58:
57:
52:
51:
46:
41:
32:
31:
19:
1860:November 11,
1858:. Retrieved
1839:. Retrieved
1833:
1817:. Retrieved
1781:
1777:
1753:
1694:
1690:
1679:
1668:. Retrieved
1656:
1646:
1627:
1621:
1602:
1596:
1587:
1578:
1567:. Retrieved
1563:
1553:
1541:. Retrieved
1539:. Ziff Davis
1530:
1487:
1483:
1477:
1432:
1426:
1414:. Retrieved
1400:
1388:. Retrieved
1377:
1365:. Retrieved
1361:the original
1351:
1342:
1333:
1322:. Retrieved
1318:
1308:
1259:
1255:
1244:
1233:. Retrieved
1229:
1219:
1208:. Retrieved
1206:. 2018-06-19
1203:
1194:
1159:
1134:. Retrieved
1129:
1104:. Retrieved
1102:. 2016-12-19
1099:
1090:
1076:
1041:
1037:
1036:computing".
1030:
1016:
1004:
963:
959:
952:
938:
929:
920:
909:. Retrieved
907:. 2016-12-19
904:
856:
849:
838:. Retrieved
832:
808:. Retrieved
804:
795:
783:. Retrieved
753:
742:
717:
713:
646:
644:
634:
621:
617:adding to it
612:
573:
565:
554:
550:
547:
544:
540:
531:
503:
488:
465:
444:
392:
384:
364:
349:
342:
323:
289:produced by
281:neuromorphic
278:
228:
221:
216:
212:
192:
190:
175:
157:
151:January 2017
148:
133:by removing
129:Please help
121:
93:
85:
61:
54:
48:
47:Please help
44:
1543:22 December
1416:22 December
1390:22 December
1367:October 17,
810:21 December
785:21 December
685:Turing test
1873:Categories
1841:2022-08-21
1819:13 January
1784:(2): 210.
1670:2022-09-28
1569:2023-10-26
1497:1906.07067
1442:1903.02504
1324:2023-10-31
1235:2023-10-30
1210:2021-08-21
1136:2021-08-21
1106:2022-09-28
1051:1907.00957
911:2021-08-21
840:2023-10-30
696:References
537:Hala Point
514:explosives
330:milliwatts
131:improve it
50:improve it
1879:Computers
1798:1530-9282
1711:1083-7159
1665:0362-4331
1522:189928531
1300:264306410
1284:0036-8075
1230:TechRadar
1068:195767210
645:In 2021,
596:Criticism
575:SpiNNaker
570:SpiNNaker
518:narcotics
480:TrueNorth
403:ResNet-50
338:Skyrmions
243:TrueNorth
56:talk page
18:TrueNorth
1814:56814452
1778:Leonardo
1729:30446581
1514:38650843
1469:70349899
1386:. TOP500
1292:37856600
1186:29335047
996:12706847
988:25104385
805:Engadget
780:18690998
654:See also
624:May 2023
583:manycore
476:trillion
472:synapses
447:seamount
424:and 256
381:Research
310:synapses
255:Qualcomm
96:May 2023
1806:1578563
1720:6656515
1564:HPCwire
1264:Bibcode
1256:Science
968:Bibcode
960:Science
722:Bibcode
516:, find
407:Yolo-v4
347:model.
334:SyNAPSE
320:Details
306:neurons
273:SyNAPSE
86:updated
1812:
1804:
1796:
1761:
1727:
1717:
1709:
1663:
1634:
1609:
1520:
1512:
1467:
1457:
1298:
1290:
1282:
1184:
1174:
1066:
994:
986:
882:
872:
778:
768:
640:Watson
450:LĆÊ»ihi
231:Watson
1810:S2CID
1802:JSTOR
1518:S2CID
1492:arXiv
1465:S2CID
1437:arXiv
1410:(PDF)
1343:ZDNET
1296:S2CID
1182:S2CID
1064:S2CID
1046:arXiv
992:S2CID
880:JSTOR
776:S2CID
637:IBM's
352:clock
302:cores
270:DARPA
251:Intel
1862:2022
1821:2017
1794:ISSN
1759:ISBN
1725:PMID
1707:ISSN
1661:ISSN
1632:ISBN
1607:ISBN
1545:2023
1510:PMID
1455:ISBN
1418:2023
1392:2023
1369:2017
1288:PMID
1280:ISSN
1172:ISBN
1009:IEEE
984:PMID
870:ISBN
834:IEEE
812:2021
787:2021
766:ISBN
414:GPUs
284:CMOS
237:and
215:and
199:and
1786:doi
1715:PMC
1699:doi
1502:doi
1447:doi
1272:doi
1260:382
1164:doi
1056:doi
976:doi
964:345
862:doi
758:doi
730:doi
718:224
619:.
557:owl
524:".
484:USB
456:or
430:MHz
422:RAM
405:or
291:IBM
1875::
1853:.
1832:.
1808:.
1800:.
1792:.
1782:21
1780:.
1776:.
1723:.
1713:.
1705:.
1695:24
1693:.
1689:.
1659:.
1655:.
1586:.
1562:.
1516:.
1508:.
1500:.
1486:.
1463:.
1453:.
1445:.
1341:.
1317:.
1294:.
1286:.
1278:.
1270:.
1258:.
1254:.
1228:.
1202:.
1180:.
1170:.
1158:.
1145:^
1128:.
1115:^
1098:.
1062:.
1054:.
1040:.
990:.
982:.
974:.
962:.
928:.
903:.
892:^
878:.
868:.
831:.
820:^
803:.
774:.
764:.
752:.
728:.
716:.
704:^
592:.
581:,
563:.
501:.
226:.
219:.
191:A
59:.
1864:.
1844:.
1823:.
1788::
1767:.
1731:.
1701::
1673:.
1640:.
1615:.
1590:.
1572:.
1547:.
1524:.
1504::
1494::
1488:2
1471:.
1449::
1439::
1420:.
1394:.
1371:.
1345:.
1327:.
1302:.
1274::
1266::
1238:.
1213:.
1188:.
1166::
1139:.
1109:.
1070:.
1058::
1048::
1042:3
998:.
978::
970::
932:.
914:.
886:.
864::
843:.
814:.
789:.
760::
736:.
732::
724::
626:)
622:(
182:)
176:(
164:)
158:(
153:)
149:(
145:.
127:.
98:)
94:(
88:.
66:)
62:(
20:)
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.