IBM and the U.S. Department of Energy’s Oak Ridge National Laboratory recently unveiled the world’s smartest, most powerful supercomputer, called Summit. At peak performance it can run 200 quadrillion calculations per second, or, in other words, if every person on Earth completed one calculation per second, it would take the world population 305 days to do what Summit can do in 1 second. Summit was built to meet the data-centric demands of the AI era, which can optimize the amount of simulation that needs to be done for faster results. Summit was designed with over 30 applications in mind, and it’s unveiling helps bring the computer industry even closer to supercomputing at scale.
Happy birthday to IBM’s first Personal Computer! The IBM 5150 PC changed the way the world works upon its debut thirty-seven years ago. As the 5150 grew in popularity, “PC” became a household name, with computers moving into small businesses, home offices, and living rooms. And with a release year of 1981, the IBM 5150 is technically a millennial—an 80’s baby that makes us proud to this day.
A computer inspired by the human brain…is it possible? IBM is part of a team helping to make this ambitious goal a reality. Together with EPFL (Swiss Federal Institute of Technology) and the New Jersey Institute of Technology, IBM Research has developed a new type of computing architecture inspired by the brain, using a technology known as a “memristive device.” These chips increase the precision associated with synaptic operations without increasing power, helping the computer process complex calculations at a level that approaches the brain’s own synaptic architecture—all without drawing too much power and overheating. There’s still a lot of progress to be made, but this research is a meaningful step towards a long-term goal: computers whose processing power could give our own neurons and synapses a run for their money.
Quantum computers make a weird and wonderful noise. Why? The answer is a cool one—or, at least, it lies in the computer’s cooling process. To help keep IBM Q quantum computers running perfectly, they are kept at the ultra-cold temperature of 15 degrees milliKelvin. This process protects our qubits, but also generates a very distinct, pleasant whooh-ing tone – the sounds of IBM Q.
Invigorating the idea of computers based on fluids instead of silicon, researchers at the National Institute of Standards and Technology (NIST) have shown how computational logic operations could be performed in a liquid medium by simulating the trapping of ions (charged atoms) in graphene (a sheet of carbon atoms) floating in saline solution. The scheme might also be used in applications such as water filtration, energy storage or sensor technology.
The idea of using a liquid medium for computing has been around for decades, and various approaches have been proposed. Among its potential advantages, this approach would require very little material and its soft components could conform to custom shapes in, for example, the human body.
NIST’s ion-based transistor and logic operations are simpler in concept than earlier proposals. The new simulations show that a special film immersed in liquid can act like a solid silicon-based semiconductor. For example, the material can act like a transistor, the switch that carries out digital logic operations in a computer. The film can be switched on and off by tuning voltage levels like those induced by salt concentrations in biological systems.
Chameleon-Like Material Spiked With Boron Comes Closer To Mimicking Brain Cells
In a new study, Texas A&M researchers in the Department of Materials Science and Engineering describe a new material that comes close to mimicking how brain cells perform computations.
Each waking moment, our brain processes a massive amount of data to make sense of the outside world. By imitating the way the human brain solves everyday problems, neuromorphic systems have tremendous potential to revolutionize big data analysis and pattern recognition problems that are a struggle for current digital technologies.
But for artificial systems to be more brain-like, they need to replicate how nerve cells communicate at their terminals, called the synapses.
Current silicon-based computing technology is energy-inefficient. Information and communications technology is projected to use over 20% of global electricity production by 2030. So finding ways to decarbonise technology is an obvious target for energy savings. Professor Paolo Radaelli from Oxford’s Department of Physics, working with Diamond Light Source, the U.K.“s national synchrotron, has been leading research into more efficient alternatives to silicon. His group’s surprising findings are published in Nature in an article titled "Antiferromagnetic half-skyrmions and bimerons at room temperature.” Some of the antiferromagnetic textures they have found could emerge as prime candidates for low-energy antiferromagnetic spintronics at room temperature.
Researchers have been working for a long time on alternative technologies to silicon. Oxides of common metals such as iron and copper are natural targets because they are already a technology staple, present in silicon-based computers, meaning there is a high chance of compatibility between the two technologies. Although oxides are great for storing information, they are not good at moving information around—a necessity for computation. However, one property of oxides that has emerged is that many are magnetic, which means it might be possible to move magnetic bits around, both in oxides and in other magnets, with very little energy required.
Scientists have succeeded in observing the first long-distance transfer of information in a magnetic group of materials known as antiferromagnets. These materials make it possible to achieve computing speeds much faster than existing devices. Conventional devices using current technologies have the unwelcome side effect of getting hot and being limited in speed. This is slowing down the progress of information technology.
[…]
The emerging field of magnon spintronics aims to use insulating magnets capable of carrying magnetic waves, known as magnons, to help solve these problems. Magnon waves are able to carry information without the disadvantage of the production of excess heat. Physicists at Johannes Gutenberg University Mainz (JGU) in Germany, in cooperation with theorists from Utrecht University in the Netherlands and the Center for Quantum Spintronics (QuSpin) at the Norwegian University of Science and Technology (NTNU) in Norway, demonstrated that antiferromagnetic iron oxide, which is the main component of rust, is a cheap and promising material to transport information with low excess heating at increased speeds. Their study has been published recently in the scientific journal Nature.
Computers may be growing smaller and more powerful, but they require a great deal of energy to operate. The total amount of energy the U.S. dedicates to computing has risen dramatically over the last decade and is quickly approaching that of other major sectors, like transportation.
In a study published online this week the journal Nature, University of California, Berkeley, engineers describe a major breakthrough in the design of a component of transistors—the tiny electrical switches that form the building blocks of computers—that could significantly reduce their energy consumption without sacrificing speed, size or performance. The component, called the gate oxide, plays a key role in switching the transistor on and off.
“We have been able to show that our gate-oxide technology is better than commercially available transistors: What the trillion-dollar semiconductor industry can do today—we can essentially beat them,” said study senior author Sayeef Salahuddin, the TSMC Distinguished professor of Electrical Engineering and Computer Sciences at UC Berkeley.
Imagine windows that can easily transform into mirrors, or super high-speed computers that run not on electrons but light. These are just some of the potential applications that could one day emerge from optical engineering, the practice of using lasers to rapidly and temporarily change the properties of materials.
“These tools could let you transform the electronic properties of materials at the flick of a light switch,” says Caltech Professor of Physics David Hsieh. “But the technologies have been limited by the problem of the lasers creating too much heat in the materials.”
In a new study in Nature, Hsieh and his team, including lead author and graduate student Junyi Shan, report success at using lasers to dramatically sculpt the properties of materials without the production of any excess damaging heat.
“The lasers required for these experiments are very powerful so it’s hard to not heat up and damage the materials,” says Shan. “On the one hand, we want the material to be subjected to very intense laser light. On the other hand, we don’t want the material to absorb any of that light at all.” To get around this the team found a “sweet spot,” Shan says, where the frequency of the laser is fine-tuned in such a way to markedly change the material’s properties without imparting any unwanted heat.
After many decades of astonishing developments, advances in semiconductor-based computing are beginning to slow as transistors reach their physical limits in size and speed. However, the requirements for computing continue to grow, especially in artificial intelligence, where neural networks often have several millions of parameters. One solution to this problem is reservoir computing, and a team of researchers led by Osaka University, with colleagues from the University of Tokyo and Hokkaido University, have developed a simple system based on electrochemical reactions in Faradic current that they believe will jump-start developments in this field.
Reservoir computing is a relatively recent idea in computing. Instead of traditional binary programs run on semiconductor chips, the reactions of a nonlinear dynamical system—the reservoir—are used to perform much of the calculation. Various nonlinear dynamical systems from quantum processes to optical laser components have been considered as reservoirs. In this study, the researchers looked at the ionic conductance of electrochemical solutions.
A collaboration between the lab of Judy Cha, the Carol and Douglas Melamed Assistant Professor of Mechanical Engineering & Materials Science, and IBM’s Watson Research Center could help make a potentially revolutionary technology more viable for manufacturing.
Phase-change memory (PCM) devices have in recent years emerged as a game-changing alternative to computer random-access memory. Using heat to transform the states of material from amorphous to crystalline, PCM chips are fast, use much less power and have the potential to scale down to smaller chips – allowing the trajectory for smaller, more powerful computing to continue. However, manufacturing PCM devices on a large scale with consistent quality and long endurance has been a challenge.
“Everybody’s trying to figure that out, and we want to understand the phase change behavior precisely,” said Yujun Xie, a PhD candidate in Cha’s lab and lead author of the study. “That’s one of the biggest challenges for phase-change memory.”
Easily printable, organic thin films can retain data for more than 10 years without power, work with low voltages – and become the building block of future computers that mimic the human brain
The internet of things is coming, that much we know. But still it won’t; not until we have components and chips that can handle the explosion of data that comes with IoT. In 2020, there will already be 50 billion industrial internet sensors in place all around us. A single autonomous device – a smart watch, a cleaning robot, or a driverless car – can produce gigabytes of data each day, whereas an airbus may have over 10,000 sensors in one wing alone.
Two hurdles need to be overcome. First, current transistors in computer chips must be miniaturized to the size of only few nanometres; the problem is they won’t work anymore then. Second, analysing and storing unprecedented amounts of data will require equally huge amounts of energy. Sayani Majumdar, Academy Fellow at Aalto University, along with her colleagues, is designing technology to tackle both issues.
Majumdar has with her colleagues designed and fabricated the basic building blocks of future components in what are called “neuromorphic” computers inspired by the human brain. It’s a field of research on which the largest ICT companies in the world and also the EU are investing heavily. Still, no one has yet come up with a nano-scale hardware architecture that could be scaled to industrial manufacture and use.
The work advances the possibility of applying quantum mechanical principles to existing optical fiber networks for secure communications and geographically distributed quantum computation. Prof. David Awschalom and his 13 co-authors announced their discovery in the June 23 issue of Physical Review X.
“Silicon carbide is currently used to build a wide variety of classical electronic devices today,” said Awschalom, the Liew Family Professor in Molecular Engineering at UChicago and a senior scientist at Argonne National Laboratory. “All of the processing protocols are in place to fabricate small quantum devices out of this material. These results offer a pathway for bringing quantum physics into the technological world.”
The findings are partly based on theoretical models of the materials performed by Awschalom’s co-authors at the Hungarian Academy of Sciences in Budapest. Another research group in Sweden’s Linköping University grew much of the silicon carbide material that Awschalom’s team tested in experiments at UChicago. And another team at the National Institutes for Quantum and Radiological Science and Technology in Japan helped the UChicago researchers make quantum defects in the materials by irradiating them with electron beams.
[It] demonstrates for the first time the possibility of performing high-fidelity excited state calculations of complex materials at very large scales within minutes on current HPC systems, paving the way for future efficient HPC software development in materials, physical, chemical, and engineering sciences.
12 tech gifts for the geekiest people in your life
There’s one in every family: The uber geek who pretty much has it all when it comes to tech basics. They don’t need a laptop or a fancy TV. But don’t worry, there’s always some gap in their collection of gadgets that you can easily fill. And remember, the more niche the product, the better gift it is for these folks.
You could get them started on building out their smart home with a lock like AugustorKevo. Or maybe they need a central hub to control their sprawling network of connected goods from, like an Amazon Echo Dot. If the nerd in your life is too young for a smart home, you can always pick them up a high-tech take on the paper airplane or a box of tinker toys likeLittleBits, to get them start on their journey to become the next great inventor.
For our full list of recommendations in all categories, don’t forget to stop by our main Holiday Gift Guidehub.
7.23andMe, gene testing company, wins back the right to tell you your risk factors for certain diseases rather than just your heritage (likely leading to wild misinterpretation by most consumers)
“Unlike most animal species, whose genomes are riddled with millions of years of mutations that have helped them adapt to a volatile world, cephalopod adaption appears to have been more a result of RNA editing. Heavy reliance on RNA editing, however it first evolved, practically would have guaranteed the need for cephalopod DNA to remain fairly stable over millennia.”
17. Just create an antibody for the bacterial-toxin associated with acne and blamo, acne vaccine…maybe
18. The ESO just discovered this firework display created by stars colliding
So the other night during D&D, I had the sudden thoughts that:
1) Binary files are 1s and 0s
2) Knitting has knit stitches and purl stitches
You could represent binary data in knitting, as a pattern of knits and purls…
You can knit Doom.
However, after crunching some more numbers:
The compressed Doom installer binary is 2.93 MB. Assuming you are using sock weight yarn, with 7 stitches per inch, results in knitted doom being…
3322 square feet
Factoring it out…302 people, each knitting a relatively reasonable 11 square feet, could knit Doom.
Hi fun fact!!
The idea of a “binary code” was originally developed in the textile industry in pretty much this exact form. Remember punch cards? Probably not! They were a precursor to the floppy disc, and were used to store information in the same sort of binary code that we still use:
Here’s Mary Jackson (c.late 1950s) at a computer. If you look closely in the yellow box, you’ll see a stack of blank punch cards that she will use to store her calculations.
This is what a card might look like once punched. Note that the written numbers on the card are for human reference, and not understood by the computer.
But what does it have to do with textiles? Almost exactly what OP suggested. Now even though machine knitting is old as balls, I feel that there are few people outside of the industry or craft communities who have ever seen a knitting machine.
Here’s a flatbed knitting machine (as opposed to a round or tube machine), which honestly looks pretty damn similar to the ones that were first invented in the sixteenth century, and here’s a nice little diagram explaining how it works:
But what if you don’t just want a plain stocking stitch sweater? What if you want a multi-color design, or lace, or the like? You can quite easily add in another color and integrate it into your design, but for, say, a consistent intarsia (two-color repeating pattern), human error is too likely. Plus, it takes too long for a knitter in an industrial setting. This is where the binary comes in!
Here’s an intarsia swatch I made in my knitwear class last year. As you can see, the front of the swatch is the inverse of the back. When knitting this, I put a punch card in the reader,
and as you can see, the holes (or 0′s) told the machine not to knit the ground color (1′s) and the machine was set up in such a way that the second color would come through when the first color was told not to knit.
tl;drthe textiles industry is more important than people give it credit for, and I would suggest using a machine if you were going to try to knit almost 3 megabytes of information.
It goes beyond this. Every computer out there has memory. The kind of memory you might call RAM. The earliest kind of memory was magnetic core memory. It looked like this:
Wires going through magnets. This is how all of the important early digital computers stored information temporarily. Each magnetic core could store a single bit - a 0 or a 1. Here’s a picture of a variation of this, called rope core memory, from one NASA’s Apollo guidance computers:
You may think this looks incredibly handmade, and that’s because it is. But these are also extreme close-ups. Here’s the scale of the individual cores:
The only people who had the skills necessary to thread all of these cores precisely enough were textile and garment workers. Little old ladies would literally thread the wires by hand.
And thanks to them, we were able to land on the moon. This is also why memory in early computers was so expensive. It had to be hand-crafted, and took a lot of time.
(little old ladies sewed the space suits, too)
Fun fact: one nickname for it was LOL Memory, for “little old lady memory.”
I mean let’s also touch on the Jacquard Loom, if you want to get all Textiles In Sciencey. It was officially created in 1801 or 1804 depending on who you ask (although you can see it in proto-form as early as 1725) and used a literal chain of punch cards to tell the loom which warps to raise on hooks before passing the weft through. It replaced the “weaver yelling at Draw Boy” technique, in which the weaver would call to the kid manning the heddles “raise these and these, lower these!” and hope that he got it right.
With a Jacquard loom instead of painstakingly picking up every little thread by hand to weave in a pattern, which is what folks used to do for brocades in Ye Olde Times, this basically automated that. Essentially all you have to do to weave here is advance the punch cards and throw the shuttle. SO EASY.
ALSO, it’s not just “little old ladies sewed the first spacesuits,” it’s “the women from the Playtex Corp were the only ones who could sew within the tolerances needed.” Yes, THAT Playtex Corp, the one who makes bras. Bra-makers sent us to the moon.
And the cool thing with them was that they did it all WITHOUT PINS, WITHOUT SEAM RIPPING and in ONE TRY. You couldn’t use pins or re-sew seams because the spacesuits had to be airtight, so any additional holes in them were NO GOOD. They were also sewing to some STUPID tight tolerances-in our costume shop if you’re within an eighth of an inch of being on the line, you’re usually good. The Playtex ladies were working on tolerances of 1/32nd of an inch. 1/32nd.AND IN 21 LAYERS OF FABRIC.
The women who made the spacesuits were BADASSES. (and yes, I’ve tried to get Space-X to hire me more than once. They don’t seem interested these days)
This is fascinating. I knew there was a correlation between binary and weaving but this just takes it to a whole nother level.
I’m in Venice, Italy several times a year (lucky me!) and last year I went on a private tour of the Luigi Bevilacqua factory.
Founded in 1875, they still use their original jacquard looms to hand make velvet.
Here are the looms:
Here are the punch cards:
Some of these looms take up to 1600 spools. That is necessary to make their many different patterns.
Here are some patterns:
How many punchcards per pattern?
This many:
Modern computing owes its very life to textiles - And to women. From antiquity weaving has been the domain of women. Sure, we remember Ada Lovelace and Hedy Lamarr, but while Joseph Marie Jacquard gets all the credit for his loom, the operators and designers were for the most part women.
