search.noResults

search.searching

dataCollection.invalidEmail
note.createNoteMessage

search.noResults

search.searching

orderForm.title

orderForm.productCode
orderForm.description
orderForm.quantity
orderForm.itemPrice
orderForm.price
orderForm.totalPrice
orderForm.deliveryDetails.billingAddress
orderForm.deliveryDetails.deliveryAddress
orderForm.noItems
Page 8


www.us- tech.com


May, 2017


Speech Recognition: Less Power, More Applications


Continued from page 1


Printed Circuit Boards from Prototype to Production


energy consumption compared to performing this operation in the cloud.” Price, Chandrakasan, and Jim


Glass, a senior research scientist at MIT’s Computer Science and Artifi- cial Intelligence Laboratory, de- scribed the new chip in a paper that Price presented at the International Solid-State Circuits Conference.


The Sleeper Wakes Today’s best-performing speech


Quick turns from 8 hours On time or FREE


FREE Laser SMT Stencil with every prototype order


Instant online quotes: www.pcb-pool.com


4777 Sunrise Blvd. Suite G Fair Oaks, CA, USA Phone: (916) 241-9062


recognizers are, like many other state-of-the-art artificial intelligence systems, based on neural networks, virtual networks of simple informa- tion processors roughly modeled on the human brain. Much of the new chip’s circuitry is concerned with im- plementing speech-recognition net- works as efficiently as possible. But even the most power-effi-


cient speech recognition system would quickly drain a device’s battery if it ran without interruption. So the chip also includes a simpler “voice activity detection” circuit that monitors ambi- ent noise to determine whether it might be speech. If the answer is yes, the chip fires up the larger, more com- plex speech-recognition circuit. For experimental purposes, the


See at SMT Hybrid Packaging, Booth 5-434


researchers’ chip had three different voice-activity-detection circuits, with different degrees of complexity and, consequently, different power de- mands. Which circuit is most power- efficient depends on context, but in tests simulating a wide range of con- ditions, the most complex of the three circuits led to the greatest power sav- ings for the system as a whole. Even though it consumed almost three times as much power as the simplest


circuit, it generated far fewer false positives; the simpler circuits often chewed through their energy savings by spuriously activating the rest of the chip.


Bandwidth Management A node in the middle of a neural


network might receive data from a dozen other nodes and transmit data to another dozen. Each of those two dozen connections has an associated “weight,” a number that indicates how prominently data sent across it should factor into the receiving node’s com- putations. The first step in minimiz- ing the new chip’s memory bandwidth is to compress the weights associated with each node. The data are decom- pressed only after they are brought on-chip. The chip also exploits the fact


that, with speech recognition, wave upon wave of data must pass through the network. The incoming audio sig- nal is split up into 10 millisecond in- crements, each of which must be evaluated separately. The MIT re- searchers’ chip brings in a single node of the neural network at a time, but it passes the data from 32 consec- utive 10 millisecond increments through it. If a node has a dozen outputs,


then the 32 passes result in 384 out- put values, which the chip stores lo- cally. Each of those must be coupled with 11 other values when fed to the next layer of nodes, and so on. So the chip ends up requiring a sizable on- board memory circuit for its interme- diate computations. But, it fetches only one compressed node from off- chip memory at a time, keeping its power requirements low. Source: www.news.mit.edu r


Silk Sensor Could Speed New Material Development


Continued from page1


Although the change was not visible to the naked eye, a red laser and a microscope built and designed by NIST were used to take photos inside the composite, showing even the most minute breaks and fissures in its interior, and revealing points where the fiber had fractured. The results were published in the journal Advanced Materials Interfaces. The materials used in the de-


sign of composites are diverse. In na- ture, composites such as crab shell or elephant tusk (bone) are made of pro- teins and polysaccharides. In this study, epoxy was combined with silk filaments prepared by professor Fritz Vollrath’s group at Oxford Universi- ty using Bombyx mori silkworms. Fiber-reinforced polymer com-


posites such as the one used in this study combine the most beneficial as- pects of the main components — the strength of the fiber and the tough- ness of the polymer. What all com- posites have in common, though, is the presence of an interface where the components meet. The resilience of that interface is critical to a com- posite’s ability to withstand damage. Interfaces that are thin but flexible are often favored by designers and manufacturers, but it is very chal-


See at SMT Hybrid Packaging, Booth 4-106


lenging to measure the interfacial properties in a composite. “There have long been ways to


measure the macroscopic properties of composites,” says researcher Jef- frey Gilman, who led the team at NIST. “But for decades the challenge has been to determine what was hap- pening inside, at the interface.”


Shown under a black light, silk samples are used to


detect damage in composites. One option is optical imaging.


However, conventional methods for optical imaging are only able to record images at scales as small as 200 to 400 nm. Some interfaces are only 10 to 100 nm in thickness, mak- ing such techniques somewhat inef- fective. By installing the RS probe at the interface, the researchers were able to “see” damage exclusively at the interface using optical mi- croscopy. Source: www.nist.gov r


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44  |  Page 45  |  Page 46  |  Page 47  |  Page 48  |  Page 49  |  Page 50  |  Page 51  |  Page 52  |  Page 53  |  Page 54  |  Page 55  |  Page 56  |  Page 57  |  Page 58  |  Page 59  |  Page 60  |  Page 61  |  Page 62  |  Page 63  |  Page 64  |  Page 65  |  Page 66  |  Page 67  |  Page 68  |  Page 69  |  Page 70  |  Page 71  |  Page 72  |  Page 73  |  Page 74  |  Page 75  |  Page 76  |  Page 77  |  Page 78  |  Page 79  |  Page 80  |  Page 81  |  Page 82  |  Page 83  |  Page 84  |  Page 85  |  Page 86  |  Page 87  |  Page 88  |  Page 89  |  Page 90  |  Page 91  |  Page 92  |  Page 93  |  Page 94  |  Page 95  |  Page 96  |  Page 97  |  Page 98  |  Page 99  |  Page 100  |  Page 101  |  Page 102  |  Page 103  |  Page 104  |  Page 105  |  Page 106  |  Page 107  |  Page 108  |  Page 109  |  Page 110  |  Page 111  |  Page 112  |  Page 113  |  Page 114  |  Page 115  |  Page 116