New Self-Learning Chip Takes Music Lessons

Before sitting down to write code for self-driving cars, software engineers have to guide them through vast amounts of information on things like traffic lights, highway lines, and other vehicles.

“But there’s an entire other set of problems where it is extremely hard to figure out what exactly you want to learn,” said Praveen Raghavan, who leads machine learning chip development at Belgium’s Imec, the advanced electronics researcher center, in a recent interview.

That is one reason why Raghavan and his colleagues built a self-learning chip to compose music, learning the craft’s hidden rules through observation rather than explicit programming. The prototype is built to faintly resemble the human brain, detecting patterns without code and using those patterns to create its own tunes.

The chip is still a work-in-progress, but it represents the latest effort in the field of neuromorphic computing. These chips aim to mimic the brain’s ability to save and apply information, consuming so little power that sensors and smartphones can run complex software for image recognition and text translation without outsourcing work to cloud servers.

The researchers, who showed an early demo this month at the Imec Technology Forum, fed the hardware pieces of classical flute music. The device detected underlying patterns in the notes, training itself to the point where it composed a new 30-second song – a simple and unpolished tune, not without a few dissonant notes.

The chip contains arrays of oxide-based random-access memory or OxRAM memory cells, which swaps between states of electrical resistance to represent bits. Its power consumption is limited because it stores and processes data simultaneously, cutting out the exhausting task of retrieving data from memory.

In training, the neuromorphic chip learns to associate notes that it hears back-to-back in multiple pieces of music. After every encounter with that pair of notes, the conductance between the memory cells storing the individual notes increases. In that way, the pattern is burned into the chip’s circuitry.

The learning process bears a resemblance to how the human brain forges connections between neurons, creating synaptic highways for familiar ideas. With training, the chip becomes more likely to play two notes together when making music. The chip also grasps larger patterns thanks to its hierarchy of memory cells.

Raghavan declined to share many specifics about the architecture or its power consumption, since the chip’s blueprints are pending patent.

Imec suggested that the chip could be used in personalized wearables, detecting heart rate changes in an individual that betray heart abnormalities. It could also let smartphones translate text or grasp voice commands without consulting the cloud, where companies Nvidia and Google are targeting number-crunching chips for deep learning.

With the rise of custom silicon for data centers, the fate of neuromorphic chips for sensors and smartphones is still uncertain. For the most part, chip designers still color within the lines drawn in 1945 by the mathematician John von Neumann, who outlined an architecture where processors always fetch data from memory.

For decades, chip designers have tried to sidestep the von Neumann architecture. Perhaps the most famous is Carver Mead, a professor of electrical engineering at Caltech, whose research into neuromorphic chips in the eighties helped create a counter-culture inspired by the human brain.

Though most neuromorphic chips have fallen short of revolutionizing computers, many companies have not given up. IBM, for instance, cast billions of silicon transistors as biological neurons in its True North chip, which the company is aiming at deep learning software and nuclear simulations in servers.

Others companies like Qualcomm are not only inspired by the human brain but are literally trying to recreate it in silicon. These spiking neural network chips, Qualcomm says, are more efficient handlers of artificial intelligence software with billions of virtual neurons. Potentially, the chips could be used from drones to servers.

Optimizing hardware for specific algorithms – and vice-versa – is vital for making more powerful chips, Raghavan said. The Imec researchers also devised a prototype chip based on magnetic random access memory and hardwired for software using deep neural networks.

“What we truly need to get the most out of the technology is for algorithms to map very efficiently to the architecture, and the architecture to map extremely well to the circuit design,” Raghavan added. “We are trying to reach across the board.”

That has been a tall order for the entire semiconductor industry, as companies struggle to kick their habit for traditional graphics and multicore processors. “It’s like getting someone who only speaks Chinese to understand a person that only knows Greek,” said Raghavan.

MEMS-Based Gyroscope Leverages Brillouin Laser Wavefronts

The rotating-mass gyroscope, which lies at the heart of inertial measurement units (IMUs), has served very successfully from the 1930s to the 1970s, guiding astronauts, spacecraft, missiles, and more. It was superseded in 1970s by the ring laser gyro (RLG) and then the fiber optic gyro (FOG), both “strap down” units with no moving parts, no need for complex gimbals, and no risk of “gimbal lock.” The next evolutionary step comprises MEMS-based gyros and accelerometers—tiny, inexpensive, and low power—which are now widely used in consumer, commercial, and many critical military applications.

However, these MEMS gyros can’t yet match the performance of the best RLGs and FOGs. The latter exploit the shifting of the interference fringes that are generated between opposing optical wavefronts, analogous to changes in musical beat notes. As the optical-gyro assembly rotates, the frequencies of each optical wave changes (known as the Sagnac effect). Thus, the fringes shift as well, with the measured shift a precise indication of the magnitude of the rotational motion.

1. The MEMS resonator is the core of the laser gyro , but many additional electronic and optical components are also needed (PM = phase modulator; SBS = stimulated Brillouin scattering; PI = proportional-integral servo loop; OBPF = optical bandpass filter; VCO = voltage controlled oscillator; ESA = electronic spectrum analyzer). (Source: Journal of the Optical Society of America)

Hoping to overcome the relatively weaker performance of MEMS-based devices, yet retain their benefits, a team at the California Institute of Technology (Caltech) is building a gyro that exploits multiple advanced electro-optical effects within a sophisticated MEMS structure (see “Microresonator Brillouin gyroscope”). Their design uses a pair of solid-state Brillouin laser wavefronts launched on counter-rotatating paths, as is the case with other optical gyros (a Brillouin wave is the result of nonlinear interaction between an optical wave and a solid material). Rather than use mirrors (as in the RLG) or optical fiber (as in the FOG), their device uses an 18-mm-diameter silica disk functioning as a precisely tuned high-Q microresonator, and it excites this resonant chamber via pumping of a single laser.

Technology Merge

While the MEMS resonator is the heart of the system, it takes a complex array of electronic, optical, and electro-optical functions to complete the system (Fig. 1), and operation involves much more than just generating that laser wavefront.

The initial Brillouin laser-pumping action induces a Stokes wave, a nonlinear periodic wave propagating along a nearly frictionless surface (analogous to an ocean wave that travels for extreme distances under the right conditions). Here, the Stokes wave pump initiates a second Brillouin wave that propagates in the opposite direction, which in turn induces yet another Stokes wave, and so on, as a Brillouin laser cascade. The overall effect thus mimics the action of the much-larger RLG/FOG, but entirely within the MEMS device.

The Caltech researchers evaluated their microresonator Brillouin gyroscope by packaging it in a small box (Fig. 2). One corner is hinged while the other was moved with an angular amplitude of 0.14 degrees at a rate of 7.5 Hz using a piezoelectric actuator.

2. The gyro prototype is housed in a box with optical-fiber input and output connectors; the 18-mm-diameter MEMS resonator is the gray device in the center. (Source: Journal of the Optical Society of America)

They reported sensitivity of 15 deg/hr/Hz2, Sagnac detection bandwidth of more than 1 kHz, and sensed a minimum rms rotation rate of 6.3 x 10-3 deg/sec (22 deg/hr), which they say is about 40 times better than previous attempts at micro-optical laser gyroscopes. They hope to improve performance by narrowing the bandwidth of the laser output.

If this project reaches practical maturity, it could takes MEMS-based guidance to another level of performance. In fact, it could very well rival the RLG, FOG, and even top-tier mechanical devices, but in a smaller, lower-power package.

Radio Shack to Close Most Retail Stores

Those of us who grew up with electronics will miss going to a Radio Shack store. After 95 years of business, Radio Shack is closing most of its retail establishments. At one time, they had over 7,000 stores—you could easily find one anywhere in the country if the need arose for a battery, cable, small parts, CB, or family radios. 

Now where do we go? Online, of course. Most of us who buy electronic parts and accessories will readily make the switch, considering the fact that so much of the electronics business in online anyway.

Over this Memorial Day weekend, Radio Shack will close over 1000 stores, leaving less than 70 corporate and 500 Radio Shack dealer stores around the country. So you have this weekend to shop one more time at your local store, if you still have one nearby.  No telling what bargains you will find at the clearance liquidation sales.  It’s probably worth a look.

If you’re interested in scoring some vintage electronics similar to this TRS-80, Radio Shack will be auctioning off items cleared out from its headquarters.

In preparation for the closing, the company cleaned out its archives at the Fort Worth headquarters. Some historic items were unearthed during the clean-out, such as original TRS-80 computers, Realistic Transistor radios, and other interesting—even classic—items. Some of these will be auctioned off in the coming month. Beginning on May 26, check in regularly over the next month at to see which of the 500 classic items will be up for auction.

The real news is that Radio Shack is not gone completely—it’s the access that’s change.  So if you’re unable to make that one last trip to the brick-and-mortar store, take a look at Radio Shack’s online version at