We’ve all got a pretty good mental image of the traditional wind-powered generator: essentially a big propeller on a stick. Some might also be familiar with vertical wind turbines, which can operate no matter which way the wind is blowing. In either case, they use some form of rotating structure to harness the wind’s energy.
But as demonstrated by [Robert Murray-Smith], it’s possible to generate electrical power from wind without any moving parts. With simple components, he shows how you can build a device capable of harnessing the wind with nothing more than vibrations. Alright, so we suppose that means the parts are technically moving, but you get the idea.
In the video after the break, [Robert] shows two different devices that operate under the same basic principle. For the first, he cuts the cone out of a standard speaker and glues a flat stick to the voice coil. As the stick moves back and forth in the wind, the coil inside of the magnet’s field and produces a measurable voltage. This proves the idea has merit and can be thrown together easily, but isn’t terribly elegant.
For the revised version, he glues a coil to a small piece of neoprene rubber, which in turn is glued to a slat taken from a Venetian blind. On the opposite side of the coil, he glues a magnet. When the blind slat starts vibrating in the wind, the oscillation of the magnet relative to the coil is enough to produce a current. It’s tiny, of course. But if you had hundreds or even thousands of these electric “blades of grass”, you could potentially build up quite a bit of energy.
Foreign hackers backed by a well-resourced government are likely to exploit a critical vulnerability in a host and VPN and firewall products sold by Palo Alto Networks, officials in the US federal government warned on Tuesday.
In worst-case scenarios, the security vendor said in a post, the flaw allows unauthorized people to log in to networks as administrators. With those privileges, attackers could install software of their choice or carry out other malicious actions that have serious consequences. The vulnerability, tracked as CVE-2020-2021, can be exploited when an authentication mechanism known as Security Assertion Markup Language is used to validate that users gave the proper permission to access a network. Attackers must also have Internet access to an affected server.
Shortly after Palo Alto Networks issued the advisory, the official Twitter account for the US Cybersecurity and Infrastructure Security Agency warned that the vulnerability is likely to be exploited in the wild by APTs, short for advanced persistent threats. APT is the term many researchers use for sophisticated hacker groups that attempt to breach select targets of interest over extended periods of time.
After years of repackaging the same basic smartwatch chip over and over again, Qualcomm has graced Wear OS with a modern smartwatch SoC. Meet the Snapdragon Wear 4100, a Qualcomm smartwatch chip that, for the first time ever, is faster than the previous chip.
The Wear 4100 uses four 1.7GHz Cortex A53 CPUs built on a 12nm manufacturing process, a major upgrade from the 28nm Cortex A7s that every other Qualcomm smartwatch chip has been up until now. It's not the state-of-the-art 7nm process that Qualcomm's high-end chip uses, and the Cortex A53 is an old CPU design, but for Qualcomm, it's a major upgrade. Between the new CPU, the Adreno 504 GPU, and faster memory, Qualcomm is promising "85% faster performance" compared to the Wear 3100.
There are actually two versions of the 4100, the vanilla "4100" and the "4100+." The plus version is specifically for smartwatches with an always-on watch face, and like previous Wear SoCs, comes with an extra low-power SoC (based around a Cortex-M0) to keep the time updated and log sensor data (like step counts). Qualcomm is promising a better display image quality in this low-power mode, with more colors and a smoother display.
Comcast is scheduled to reinstate its home-Internet data cap tomorrow, July 1, after more than three months in which customers were provided unlimited data to help them through the coronavirus pandemic. AT&T, by contrast, announced today that it is "continuing to waive home-Internet data overage charges for AT&T Internet customers through September 30."
Comcast and AT&T suspended their data caps and overage fees in mid-March, initially promising two months of unlimited data. The companies later extended that pledge to June 30, but Comcast hasn't granted any further extensions. We contacted Comcast yesterday but didn't receive answers to questions about its data cap, and Comcast's website still says the data-cap waiver only goes through June 30. While Comcast didn't answer the data-cap questions, a spokesperson pointed out that the cable company extended other pandemic offers for college students and people with low incomes beyond June 30, and it is keeping its Wi-Fi hotspots open to the public for free for the rest of 2020.
Many US states are taking steps toward reopening their economies, which might reduce usage of home-broadband networks. But the pandemic is far from over, as the CDC reported over 40,000 new daily cases in the United States each day from June 25 to June 28, including the highest-ever daily count of 44,703 on June 27.
Today's Dealmaster is headlined by an Amazon Gold Box sale on storage products from SanDisk and WD. The discounts cover a number of microSD cards, SD cards, flash drives, and portable hard drives, among other devices.
The highlights include the 128GB version of SanDisk's microSDXC Card for Nintendo Switch, a speedy and reliable microSD card we recommended in our guide to the best Nintendo Switch accessories, down to $22.35. This isn't the lowest price we've ever seen, but it's a nice drop from its usual $28 and the cheapest this model has been since it went for $20 on Black Friday. If you're more concerned with storage space than performance, the 400GB SanDisk Ultra is a slower card but offers a much lower cost-per-GB ratio for devices like the Switch. It's currently down to $48 from its usual $60.
Elsewhere, the SanDisk Extreme is a durable portable SSD we've tested and recommended before—its 1TB model is down to $136 from a typical street price around $170. The company's iXpand Flash Drive Go, meanwhile, is a useful flash drive with a Lightning connector that's about $10 off its normal going rate. There are plenty more deals beyond that, but just note that, like all Gold Box sales, the deals are available for today only.
Small waterways give life in the form of drinking and irrigation water, but can also be very destructive when flooding occurs. In the US, monitoring of these waterways is done by mainly by the USGS, with accurate but expensive monitoring stations. This means that there is a limit to how many monitoring stations can be deployed. In an effort to come up with a more cost-efficient monitoring solution, [Rohan Menon] and [Ian Vernooy] created Aquametric, a simple water level, temperature and conductivity measuring station.
The device is built around a Particle Electron that features a STM32 microcontroller and a 3G modem. An automotive ultrasonic sensors measures water level, a thermistor measures temperature and a pair of parallel aluminum plates are used to measure conductivity. All the data from the prototype is output to a live dashboard. The biggest challenges for the system came with field deployment.
The great outdoors can be rather merciless with our ideas and electronic devices. [Rohan] and [Ian] did some tests with LoRa, but quickly found that the terrain severely limited the effective range. Power was another challenge, first testing with a solar panel and lithium battery. This proved unreliable especially at temperatures near freezing, so they decided to use 18 AA batteries instead and optimized power usage.
The mounting system is still an ongoing challenge. A metal pole driven into the riverbed at a wider part ended up bent (probably from ice sheets) and covered in debris to the point that it affected water level readings. They then moved to a narrower and shallower section in the hopes of avoiding debris, but the rocky bottom prevented them from effectively driving in a pole. So the mounted the pole on a steel plate which was then packet with rock to keep it in place. This too failed when it tipped over from rising water levels, submerging the entire sensor unit. Surprisingly it survived with only a little moisture getting inside.
For the 2020 Hackaday Prize, Field Ready and Conservation X Labs have issued challenges that need require some careful consideration and testing to build things that can survive the real world. So go forth and hack!
Enlarge/ Chinese President Xi Jinping learns about the progress on a COVID-19 vaccine during his visit to the Academy of Military Medical Sciences in Beijing on March 2, 2020. (credit: Getty | Xinhua News Agency)
China has approved an experimental COVID-19 vaccine for use in its military after early clinical trial data suggested it was safe and spurred immune responses—but before larger trials that will test whether the vaccine can protect against SARS-CoV-2 infections.
China’s approval marks the first time any country has approved a candidate vaccine for military use. The country’s Central Military Commission, made the approval June 25, which will last for a year, according to a filing reported by Reuters.
The vaccine, developed by biotech company CanSino Biologics and the Chinese military, is a type of viral vector-based vaccine That is, researchers started with a viral vector, in this case a common strain of adenovirus (type-5), which typically causes mild upper respiratory infections. The researchers crippled the virus so that it doesn’t replicate in human cells and cause disease. Then, they engineered it to carry a signature feature of SARS-CoV-2—the coronavirus’ infamous spike protein, which juts out from the viral particle and allows the virus to get a hold on human cells.
You might have caught Maya Posch’s article about the first open-source ASIC tools from Google and SkyWater Technology. It envisions increased access to make custom chips — Application Specific Integrated Circuits — designed using open-source tools, and made real through existing chip fabrication facilities. My first thought? How much does it cost to tape out? That is, how do I take the design on my screen and get actual parts in my hands? I asked Google’s Tim Ansel to explain some more about the project’s goals and how I was going to get my parts.
The goals are pretty straightforward. Tim and his collaborators would like to see hardware open up in the same way software has. The model where teams of people build on each other’s work either in direct collaboration or indirectly has led to many very powerful pieces of software. Tim’s had some success getting people interested in FPGA development and helped produce open tools for doing so. Custom ASICs are the next logical step.
Who Needs Open Source ASICs?
Of course, FPGAs and ASICs aren’t the answer to every problem. We can’t help but notice that some examples you see — including ours — are sometimes better for learning than actually practical. For example, the classic sample for learning about state machines on an FPGA is a traffic light. Why not? Everyone sort of understands what it is supposed to do, it has clear state logic, and you can make it as simple as you like or quite complex if it senses vehicles and pedestrian crosswalk buttons or changes based on schedules.
However, if you were really building a traffic light, it wouldn’t make a lot of sense to do it in an FPGA. Even the simplest microcontroller would be up to the task and would be cheaper both to buy and in terms of engineering costs by a wide margin.
ASICs occupy a similar niche, but with a little bit of a difference. On the plus side, they should be denser, faster, and less power hungry than a similar FPGA. That makes sense because the ASIC is sort of an FPGA where the interconnections are made with dedicated metal lines instead of being generally configurable. You can also put down exactly the circuits you want — or, at least, choose from a variety of cells instead of having to use whatever the FPGA’s architect decided you need. You can even include analog cells alongside digital circuitry.
On the negative side, ASICs are not for the sloppy. Historically, taping out an ASIC has been very expensive. So you have a run of parts but — oops — you forgot that counter needs to reset to a non-zero number. In an FPGA, that’s a minor annoyance; you simply change the configuration — especially now that one time programmable FPGAs are rare outside of certain applications. Even if you have to trash an FPGA and program another one, they are generally not very expensive unless they are radiation hardened or very large devices.
If you make that mistake on an ASIC, you are in big trouble. You can’t change anything on the parts you have. You have to have a new batch built with new upfront costs. In the commercial world, that kind of mistake can be career-ending.
Tim makes it clear that his target audience isn’t the professional building custom ASICs, though. It is us. The hackers and tinkerers that want to create custom ICs. There may be some student market, too, although schools often have deals to make that feasible already.
Tim does point out, though, that a lot of those school deals are bound up with nondisclosure agreements the students have to sign, so it’s possible that open tools will spur new published research which would be a good thing. Still, I get the sense they think most of the interest will be from our community.
Notable about this process is that the 130 nm process being used isn’t cutting edge technology. The Skywater Technologies fab was built by Cypress Semiconductor in 1991 in Bloomington, Minnesota. Tim says professional designers have moved so far from these large geometries that our designers may have to rediscover some lost knowledge along the way to get the most from an IC made on the larger processes now. But the existing infrastructure is a big part of what makes this project more affordable.
So How Do You Get Them?
Tim had a lot to say about cell libraries that are eminent and how each one was tuned for a different purpose (e.g., high density or low power or high speed). However, we wanted to know how we’d get actual parts. Apparently, some of the details or still being worked out.
Chip scale devices on a penny by Cp82 CC-BY-SA 3.0
In November, they plan to order a multiproject wafer with 40 slots. They don’t know yet if they will have to beg and plead to get 40 designs or if they will have to winnow the select down from all possible candidates. If you are one of the 40, you’ll get about 10mm square to play with and wind up with somewhere around 100 to 300 chips in chip-scale packaging (CSP). You can see a typical CSP sitting on a US penny in the accompanying photo.
There are a few stipulations. You’ll submit your design on GitHub (or some similar public repository), so your design is going to be open source. That means even if you aren’t one of the 40, you’ve just put your chip out for the world to see. The foundry will automatically check your design to meet certain technical criteria. At this early point there doesn’t seem to be a firm plan on how they will select designs for inclusion in the first run. Presumably, if there are a lot of entrants and things work well, there will be more wafers in 2021.
There are still a lot of unanswered questions. Can you pay to get your own tape out? If so, do you still have to be open source? What if you have some made and then want more? How much does that cost? This is very early and we do we not yet know the answers to these questions, but details will come together over time.
The Key
Like I said earlier, ASICs aren’t for everyone and they certainly aren’t for people who test and debug as they go. Verification is essential for a successful ASIC project. That means a lot of this will hinge on the simulation tools available and the quality of the models available. Spending a lot of time and money getting ICs that won’t work at the speeds you need, consume more power than you expected, or simply don’t work is heartbreaking.
Many times an FPGA can be used to validate some or all of your design before trying to go to an ASIC. When that works, it works well. However, because of the differences between the two technologies, it isn’t as simple as thinking of an ASIC as a fixed FPGA. You have the same problems you might have going from a hand-wired circuit to a PCB. Logically they are the same. But we all know you can have problems with that transition because of the different characteristics. It is the same problem here. How do you test your analog cells? Will the clock distribute the same? And ASICs have speed or power requirements which are difficult to mimic in a validation stage.
Tim Ansel gave an online talk today officially announcing the project. Take a look for more details on the process node itself and the tools used to design for it:
So will you try to design your own IC? I’ve been involved in ASIC development before, but I still might be interested in doing my own personal project just to be able to do all the steps. Let us know what IC you want to design — or see someone else design — in the comments.
After eliminating temporary "hazard" pay raises, Amazon is saying "thank you" to its hourly workers with a one-time bonus of $500, while at least one Amazon warehouse has been found to have a COVID-19 rate four times higher than the general population nearby.
Amazon yesterday announced its one-time bonuses for "front-line" employees. Full-time workers in warehouses and Whole Foods stores, as well as full-time delivery drivers, will receive $500. Part-time workers in those roles will get $250, and Amazon Flex drivers who worked 10 hours or more will get $150. Managers on-site in distribution centers or Whole Foods stores will get $1,000, and owners of the third-party firms that handle delivery for Amazon will get $3,000.
The company saw a massive spike in consumer demand as in-person retail shuttered around the nation and the world this spring due to the COVID-19 pandemic. Amazon hired an additional 175,000 employees in its warehouses, logistics, and grocery businesses since early March to meet increased demand. The company also increased wages by $2 in warehouses, to a minimum of $17 per hour, to get new workers in the door.
Enlarge/ UC graduate student Brian Lane climbs out of the Perdido Reservoir. (credit: Photo/Nicholas Dunning)
For centuries, Tikal was a bustling Maya city in what is now northern Guatemala. But by the late 800s CE, its plazas and temples stood silent, surrounded by mostly abandoned farms. A recent study suggests a possible explanation for its decline: mercury and toxic algal blooms poisoned the water sources that should have carried the city through dry seasons.
Tikal’s Maya rulers built the city’s reservoirs to store water from rain and runoff during the winter months. The pavement of the large plazas in the heart of the city tilted slightly, helping funnel rainwater into the reservoirs. Over the centuries, dust and litter settled into the bottom of the reservoirs, too, providing a record of what the environment around Tikal was like—and what was washing into the city’s water supply. University of Cincinnati biologist David Lentz and his colleagues sampled layers of sediment dating back to the mid-800s, and they found that two of Tikal’s central reservoirs would have been too polluted to drink from.
An X-ray fluorescence spectrometer (which identifies the chemicals in a sample based on how they react to being zapped with an X-ray light) revealed that the sediment on the bottom of the reservoirs was laced with dangerous amounts of mercury. Lentz and his colleagues also found ancient DNA from blue-green algae, or cyanobacteria, which can produce deadly toxins.
A model shows off a prototype form factor for a holographic VR display. [credit: Facebook Research ]
For all the advancements in virtual reality technology in recent years, one major factor still holding the space back is the size and relative discomfort of current headset design. Even the most compact and comfortable VR headsets today still resemble something like a cross between ski goggles and a motorcycle helmet, requiring massive headstraps to secure a heavy display that protrudes multiple inches away from the face. Reference designs for "eyeglasses" style VR displays help a bit, but they still look like coke-bottle spectacles from a steampunk cosplay event (and provide a limited field of view, to boot).
Now, researchers at Facebook Reality Labs are using holographic film to create a prototype VR display that looks less like ski goggles and more like lightweight sunglasses. With a total thickness less than 9mm—and without significant compromises on field of view or resolution—these displays could one day make today's bulky VR headset designs completely obsolete.
In the newly published ACM Siggraph paper Holographic Optics for Thin and Lightweight Virtual Reality, researchers Andrew Maimone and Junren Wang detail the optics behind their lightweight prototype. The key to the thinness is a series of flat, polarized films that use a "pancake optics" light-folding technique to reflect the displayed image multiple times in a small space.
Enlarge/ CCTV security guard in the mall building.
Detroit's police chief admitted on Monday that facial recognition technology used by the department misidentifies suspects about 96 percent of the time. It's an eye-opening admission given that the Detroit Police Department is facing criticism for arresting a man based on a bogus match from facial recognition software.
Last week, the ACLU filed a complaint with the Detroit Police Department on behalf of Robert Williams, a Black man who was wrongfully arrested for stealing five watches worth $3,800 from a luxury retail store. Investigators first identified Williams by doing a facial recognition search with software from a company called DataWorks Plus. Under police questioning, Williams pointed out that the grainy surveillance footage obtained by police didn't actually look like him. The police lacked other evidence tying Williams to the crime, so they begrudgingly let him go.
Now Vice's Jason Koebler reports that Detroit Police Chief James Craig acknowledged the flaws with its facial recognition software at a Monday event.
In the video after the break, [Robert] shows two different devices that operate under the same basic principle. For the first, he cuts the cone out of a standard speaker and glues a flat stick to the voice coil. As the stick moves back and forth in the wind, the coil inside of the magnet’s field and produces a measurable voltage. This proves the idea has merit and can be thrown together easily, but isn’t terribly elegant.
ASICs occupy a similar niche, but with a little bit of a difference. On the plus side, they should be denser, faster, and less power hungry than a similar FPGA. That makes sense because the ASIC is sort of an FPGA where the interconnections are made with dedicated metal lines instead of being generally configurable. You can also put down exactly the circuits you want — or, at least, choose from a variety of cells instead of having to use whatever the FPGA’s architect decided you need. You can even include analog cells alongside digital circuitry.
