Intel Pushes Moore’s Law Along: 10 nm

intel logo: a light blue print of intel with a oval starting from the bottom of the "l" of the name on the right and wrapping back around to the bottom of the "l"

Moore’s Law (which states the number of transistors per square inch doubles roughly every twelve {12} to eighteen {18} months) has had repeated claims that it would end as the limits of silicon are hit and the size approaches that where quantum effects take over, yet it keeps proving the naysayers wrong. IEEE Spectrum reports that central processing unit (CPU) manufacturer Intel is pushing Moore’s law further as it plans to push out computer and mobile processors with transistors that are just ten {10} nanometers (nm) wide. However, these transistors are going to be a bit different than your average transistor…

Intel plans on, for the first time in quite a long time, decrease the size of the gate (the piece of a transistor that switches it “on” or “off”) and the gate pitch – the size of the material that exists between one gate and another. They are also planning on making two {2} improvements on their transistor design within it’s lifetime. Intel claims this change will create a processor that, while still more expensive than the last generation, will still be cheaper per-transistor than it’s previous product offerings. Though, as with the modern generations of processors, don’t expect much difference in clock speed. What about the other producers?

Intel is also planning on allowing other processor manufacturers to use their manufacturing facilities to produce their own chips. This is less likely to be an invitation to competitors and more an invitation for manufacturers of specialized processor and chipset manufacturers. Global Foundries, the manufacturer that spun off from AMD years ago, is planning on skipping ten {10} nm altogether and jumping right to seven {7} nm in 2018.

Moore’s Law is expected to end once transistors reach 5nm. Below that size the effects of quantum physics start taking over and electrons begin “tunneling” – a quantum effect where an electron suddenly tunnels through insulating material and pops out on the other side. Essentially an electron in one transistor could suddenly end up in the one next to it – a one {1} becomes a zero {0} and a zero {0} becomes a one {1} – yikes! It is yet to be seen if something – a solution or perhaps a new material – appears to continue Moore’s Law in the future.

Canon Global Advancement: Read the Side of an Airplane!

Canon's corporate logo consisting of red lettering with sharp serifs.

Canon, a leader in technology for digital cameras has made an amazing breakthrough in complementary metal-oxide semiconductor (CMOS) technology. Phys.org reports that Canon has developed a CMOS with the highest pixel density ever for a 35 mm full-frame sensor and simultaneously overcome the major hurdles to getting there. At 250 megapixels, or roughly 250 million pixels, it is one beast of a sensor!

A CMOS sensor that appears with a gradient of blue to green to orange from top left to bottom right surrounded by the circuit it is situated on and the thin metal pins the protrude from all four sides.
The Canon-developed approximately
250-megapixel CMOS sensor via Canon Global

The features of this new visual sensor are nothing short of revolutionary. Canon shows it’s new sensor as being able to perform massive digital zoom without pixelation. At 1,920 x 1,080 pixels the sensor is capable of nearly 125x digital zoom and at 3,840 x 2,160 (4K) it is still capable of roughly 30x digital zoom without loosing resolution! They claim that at such high resolution, the digital zoom is able to capture a readable photo of the lettering on the side of an airplane that is over 11 miles (18 kilometers) up! A typical commercial airliner travels at between 30,000 and 40,000 feet or between 5.5 and 7.5 miles high.

Canon also had to overcome some age-old issues with increasing pixel count. As CMOS sensor pixel count increases, so too must the amount of data sent from the sensor into the camera’s memory. Unfortunately this increase in needed bandwidth isn’t always available and you end up with delays before all the information can be stored. If the signal can’t be read all at once you can also get discrepancies where parts of the images are read at slightly different times. This means if a boat is passing through the frame at high speed it could be cut in half or show up in two places within the same image. However these were not issues with the new sensor as circuit miniaturization and an increased speed of signal processing is able to handle the larger bandwidth throughput. In fact, the new technology, even at the highest resolution the sensor is capable of (19,580 x 12,600 pixels), was still able to capture video at 5 frames per second (fps)!

Canon is currently looking at developing this sensor for surveillance equipment, possible medical applications, and manufacturing. It will be interesting to see what becomes of this and how the advances affect camera technology all the way down to cell phones over time.

Towards Affordable Transparent Aluminum

Screen capture of the molecular diagram of transparent aluminum on an old Apple computer screen from the movie Star Trek 4: The Voyage Home
[dc]C[/dc]ontrary to popular belief, transparent aluminum is real… It’s just prohibitively expensive (or at least it was). It is known in the scientific world as magnesium aluminate or spinel. It is a mineral ceramic that is capable of allowing the visible and infrared spectrum to pass through (which is why it is often used in military applications). It is significantly stronger and harder than glass.

The U.S. Navel Research Laboratory (NRL) has discovered a cheaper method of producing the material that also requires less energy. The new method discovered uses a low-temperature hot-press that limits the size of the spinel only to the size of the press used to form it. The laboratory team, lead by Dr. Jas Sanghera, has agreed to hand over the method to the commercial industry to allow businesses to fully utilize the promising material. Because of it’s previously high cost it was primarily used in military and police armor.

Application

What can this material be used for now? A wide range of things. Lets start with cell phones. A phone screen made of transparent aluminum would be very difficult to scratch and would not shatter if you dropped the phone onto concrete. Since this material is easily bullet-resistant it could also be used to lower the cost and weight of armor for vehicles. Bullet-resistant glass for vehicles used to protect high-profile individuals such as celebrities, business persons, and politicians could be reduced in thickness while still providing the same ballistics protection. Current bullet-resistant material of choice is thick Plexiglas but to prevent most bullets from penetrating the material has to be very thick and therefore very heavy. Transparent aluminum could be used instead which would reduce weight, be easier to install, and reduce the amount of fuel the vehicle used. Because of the optical properties it is also likely to be used by the solar power industry in the future as well. It could provide better protection for solar cells with the possibility of even enhancing efficiency if the optical properties could be tuned.

The video below is from the movie Star Trek IV: The Voyage Home where Chief Engineer Montgomery “Scotty” Scott (played by the late James Doohan) along with Doctor Leonard “Bones” McCoy (played by the late Jackson DeForest Kelley) have traveled back in time and are in need of a container to hold two humpback whales and the water needed for them to survive. Scotty pulls up the chemical structure of transparent aluminum on the computer for the manager of Plexicorp, a ceramics manufacturer, and offers the deal of a lifetime.

Google’s Cell Service Play

Google Project fi logo - a green and blue lower-case "f" and yellow "i". The dot above the "i" is white and overlaps the cross of the "f".

Google likes to jump into a number of businesses that involve technology. They are heavily involved in robotics and are developing a self-driving car, conduct a number of research projects, jumped into the cloud computing ring, more recently became an ISP (Internet Service Provider) by rolling out fiber-optic internet to a number of cities across the United States, and develop the Andorid OS (operating system) that runs roughly half of the world’s cell phones. Now they are looking to take over your cell phone service as well. Google just announced Project Fi, their new mobile phone service.

Google Project fi logo - a green and blue lower-case "f" and yellow "i". The dot above the "i" is white and overlaps the cross of the "f".
Google’s Project fi Logo

The new service — currently only open to a few who request an invite — offers mobile phone service for $20 per month with data starting at $30 per month for 3 GB (gigabytes) — total of $50 per month. That is a little underwhelming given that other wireless carriers offer similarly-priced plans. It is not until you add in the discounts and features they it becomes mildly intriguing. First of all they refund you for the data you did not use. So you get refunded for the amount of data you don’t use under $3. So if you only use 1 GB in a month they will refund you $20 (data is charged at $10 per GB). There are no contracts.

One of the major drawbacks of this service is the phone selection. There is none. Currently you can only use the Motorola-produced Google Nexus 6. Sorry, no Apple iPhones here.

Where this show gets somewhat more interesting is how the service works: It uses 2 networks. Google partnered with Sprint and T-Mobile — both providers use similar technology in their networks — and the phone can simply hop onto the network that has the strongest signal. This probably increases the signal strength mildly since Sprint and T-Mobile are the smaller networks operating in the U.S. The other way to make calls is over a Wi-Fi network (including the many open networks available at restaurants, coffee shops, airports, and other offices and retail stores nationwide). However, even that is not new: T-Mobile already offers a service that allows for calls over a Wi-Fi connection.

On the plus side if you travel a lot it could be a sigh of relef. Some other mobile service providers make you jump through hoops, pay a little to a lot more for service and/or data, or simply don’t offer service in other countries. This new plan from Google works in more than 120 countries (since Sprint and T-Mobile use the same wireless technology the majority of service providers outside the U.S. use it is more compatible) though data speed is limited since only 3G connections will work. They also do not charge any more for data when traveling. It’s still the same $10 per GB. International calling rate of $0.20 per minute apply. No extra charges for texting internationally.

It’s an modest start — it’s not likely to cause a mass-exodus from other cell service providers — but will be interesting to see how their service evolves.

Carbon Nanotube Filtering Breakthrough

Yellow stick of butter with two lines - one drawn across the top to represent a semiconducting wire that has not melted through the butter and another that has sunk to the bottom of the butter representing a conducting wire. The example shows a way to purify - or sort - carbon nanotubes with different properties.

Carbon nanotubes, microscopically-thin wires of carbon atoms, can be produced in sufficient quantity but not sufficient quality for electronics. They often include a bundle of wires where some are conductive, like the power wires going from your computer to the wall outlet, and some are semiconducting — the kind needed for processing information. Science Daily reports on a carbon nanotube purification breakthrough by a research team at University of Illinois at Urbana–Champaign, lead by professor John Rogers. Efforts to purify or sort conducting from semiconducting nanotubes have been expensive and require many steps. The new method discovered can be explained easily.

Imaging you have a stick of butter and you lay two thin metal wires — one a standard metal wire and the other a semiconducting wire — over top of the butter then attach the positive and negative electrodes of a battery to each end of wire. After a few seconds what happens? The conductive metal wire heats up and sinks into the butter stick. The semiconducting wire does not heat up nearly as much due to restricting the electron flow — so it does not sink into the butter as deeply. Instead the semiconducting wire stays nearly on the surface of the butter. Once the process is complete you can easily separate the conductive and semiconducting wire since the conductive one is at the bottom of the butter stick. Image edited to show example:

Yellow stick of butter with two lines - one drawn across the top to represent a semiconducting wire that has not melted through the butter and another that has sunk to the bottom of the butter representing a conducting wire. The example shows a way to purify - or sort - carbon nanotubes with different properties.

WARNING: DO NOT try the experiment described above! It is dangerous and could lead to burns or even an explosion — the battery and wires will heat up and can remain hot and the battery may even explode!

This new method of using current-induced heat to separate nanotubes of different properties is easy to do and is compatible with current manufacturing methods.

WiFi Traffic Management Algorithm

Visual representation of the 2.4 GHz WiFi frequency channels. Each channel is represented by a dotted half-circle representing 22 MHz of bandwidth. The half-circles representing the 3 front channels (1, 6, and 11) have solid outlines. The others overlap behind and between the front 3 channels except for channel 14 which only overlaps the edges of the 12th and 13th channels.

Phys.org reports on a new algorithm developed by a doctoral student at École polytechnique fédérale de Lausanne (EPFL) that changes frequencies and bandwidth usage based on the type of data packets being sent and received. Many routers today are set by default to use channel 6 of the 2.4 GHz frequency which causes a build-up of WiFi traffic on that channel. The problem is that many other channels overlap and use much of the same frequencies. In fact, while there are 14 total channels made available in the 2.4 GHz range, many countries ban the use of some of those frequencies. In the United States (US) channels 12 through 14 are not able to be used yet are the ones with the greatest frequency gap between channels. In effect, because the frequency bands overlap you can argue that there are really only 3 available spaces to transmit data in the 2.4 GHz WiFi band.

Visual representation of the 2.4 GHz WiFi frequency channels. Each channel is represented by a dotted half-circle representing 22 MHz of bandwidth. The half-circles representing the 3 front channels (1, 6, and 11) have solid outlines. The others overlap behind and between the front 3 channels except for channel 14 which only overlaps the edges of the 12th and 13th channels.
Visual representation of the 2.4 GHz WiFi frequency channels and how they overlap (22 MHz channels). Creative commons licensed image by Michael Gauthier on Wikimedia Commons.

The graph above shows the frequency channels for the 2.4 GHz WiFi range and how the channels overlap. Most routers are set to channel 6 by default and while they may change channels depending on availability they generally pick a channel and stick with it. In addition, many routers will use up to 8 of these channels at the same time. The problem is that this rather small range gets filled up in areas where many routers are being run and essentially cause a traffic jam of data. The other problem is that because routers will often stick with a set channel other may actually be open and unused.

The new algorithm would determine the bandwidth requirements of the data being sent and received and would select an appropriate channel and width. It essentially removed the idea of “channels” and instead divvies up the available frequency range into “lanes.” Some of the lanes are specialized similar to having a carpool or bike lane. As an example, if all you did was check your e-mail and browse a few websites you don’t need much bandwidth. The new algorithm would utilize a small amount of bandwidth – say within channels 1 and 2 – for just website browsing and email. Videos such as Vimeo and YouTube, which require much more bandwidth, may get a large chunk of channels 6 through 10 to use, and the remaining could be used for various other purposes such as websites with larger images, chat programs, and cell-phone updates. It spreads out the use over the available bandwidth and specialized certain areas for things like low-bandwidth data such as web and email, cell-phone updates, and high-bandwidth videos. The developer claims that it could increase typical router throughput by up to seven times (7X).