Subject: General Tech | October 13, 2014 - 08:57 PM | Scott Michaud
Tagged: processors, microprocessor, FinFET, fab
Ah, Solid State Physics. Semiconductors are heavily based on this branch, because it explains the physical (mechanical, electrical, thermal, etc.) properties of solids based on how their atoms are organized. These properties lead into how transistors function, and why.
Put it back, Allyn.
Anandtech has published a seven-page article that digs into physics and builds upon itself. It starts with a brief explanation of conductivity and what makes up the difference between a conductor, an insulator, and a semiconductor. It uses that to build a simple transistor. From there it explains logic gates, wafers, and lithography. It works up to FinFETs and then keeps going into the future. It is definitely not an article for beginners, but it can be progressed from start to finish given enough effort on the part of the reader.
While this was not mentioned in the article, at least not that I found, you can derive the number of atoms per "feature" by dividing its size by the lattice-distance of the material. For silicon, that is about half of a nanometer at room temperature. For instance, 14nm means that we are manufacturing features that are defined by less than 30 atoms (up to rounding error). The article speculates a bit about what will happen after the era of silicon. This is quite interesting to me, particularly since I did my undergraduate thesis (just an undergrad thesis) on photonic crystals, which route optical light across manufactured defects in an otherwise opaque solid to make an optical integrated circuit. It has the benefit of, with a mixture of red, orange, and maybe green lasers, being able to "go plaid".
If you are interested, be sure to read the article. It is a bit daunting, but much more manageable than most sources. Congratulations to Joshua Ho and anyone else who might have been involved.
Subject: General Tech | October 14, 2014 - 02:43 PM | Scott Michaud
Tagged: Samsung, 802.11ad, wigig
Samsung Electronics, a member of the WiGig Alliance, has just announced an implementation that is capable of achieving 4.6 gigabit (575MB/s) speeds under the 802.11ad standard. Samsung claims that they have overcome "the barriers to commercialization" of wireless over 60GHz. This band has several disadvantages, including resonance with oxygen molecules (included under the blanket of "path loss" in the press release) and its opacity to many solid objects (referred to as "weak penetration properties" in the release).
Image Credit: Wikipedia
Some features that Samsung credits themselves with are beam-forming with less than four-tenths of a millisecond latency and the ability to track multiple devices simultaneously. Beam-forming in particular is said to help offset the mostly line-of-sight properties of earlier 60GHz prototypes. This allows the signal to be directed toward devices, typically by manipulating interference patterns to reduce the energy lost by transmitting to locations without a receiver and thus giving more energy to the locations that do.
Its usage as a product will mostly depend on how tolerant they are to non line-of-sight situations. This rate is comparable to a high-end SATA SSD. Samsung claims that it will be useful for their Smart Home and Internet of Things initiatives, similar to the Stanford and Berkeley announcement last month, but also mention it in terms of medical devices.
Subject: General Tech | October 14, 2014 - 03:28 PM | Jeremy Hellstrom
Tagged: predix, Cisco, Intel, GM, verizon, Privacy, security
GM's Predix asset management platform has been used for a while now, after they came to the realization that they were in the top 20 of the largest software developers on the planet. They found that by networking the machines in their factories as well as products that have been shipped to customers and are seeing active use that they could increase the efficiency of their factories and their products. They were aiming for 1% increase, which when you consider the scale of these industries can equate to billions of dollars and in many cases they did see what they had hoped for.
Now Cisco and Intel have signed up to use the Predix platform for the same results, however they will be applying it to the Cloud and edge devices as well as the routers and switches Cisco specializes in. This should at the very least enhance the ability to monitor network traffic, predict resource shortages and handle outages with a very good possibility of a small increase in performance and efficiency across the board. This is good news to those who currently deal with the cloud but it is perhaps worth noting that you will be offering up your companies metrics to Predix and you should be aware of any possible security concerns that may raise because of that integration to another system. You could however argue that once you have moved to the cloud that this is already happening.
"GE, Intel, Cisco, and Verizon have announced a big data deal to connect Predix — GE’s software platform — to machines, systems, and edge devices regardless of manufacturer."
Here is some more Tech News from around the web:
- Flexible FinFETs work at high temperatures @ Nanotechweb
- Firefox 33 Arrives With OpenH264 Support @ Slashdot
- Intel 'underestimates error bounds by 1.3 QUINTILLION' @ The Register
- Linux Foundation announces Dronecode alliance for open source Drone ware @ The Inquirer
- NETGEAR AC750 WiFi Extender @ HardwareHeaven
- Apotop Wi-Copy @ Phoronix
Subject: Cases and Cooling | October 19, 2014 - 01:52 PM | Scott Michaud
Tagged: heatsink, fanless, air cooling
There are many interesting ways to pull heat away from a processor. You can submerge your device in mineral oil or even phase-change fluid (such as "Novec"). You can push cool fluid up to the thing that you are trying to remove heat from and then pump it away through a radiator. If using air, you can make use of vapor chambers and the convection current formed as devices heat up. The goal is to abuse one or more interesting material properties to store energy and move it somewhere else.
Image Credit: HT4U.net
Or you can just have an obscene, gigantic mass of metal with more fins than the NHL. According to FanlessTech, these are three heatsinks that are not yet available (and may never be). Two of them have three towers, connected to the base by heat pipes, and the last one has four.
Image Credit: ExtraHardware.cz
Personally, I would be a bit uncomfortable about buying a PC like that unless I needed absolutely silent or top air cooling performance. The amount that it hangs over RAM or nuzzles against add-in boards seems sketchy to me, especially if you need to swap a DIMM or two at some point, but I always use stock coolers at reference voltage and frequency so what do I know?
Image Credit: PConline.com.cn
Yes, that would be a regular, ATX motherboard.
When will these prototypes become available? Who knows if they even will. Still, if you have a need for cooling solutions that are a little over-the-top, you might be able to get your hands on these some day. There's nothing wrong with adding more mass and surface area, rather than doing something fancy. It works, and it probably works really well.
Subject: General Tech | October 19, 2014 - 08:20 PM | Scott Michaud
Tagged: unreal engine 4, epic games
Epic Games released Unreal Engine 4.5 last Tuesday, and it is one of their largest releases since launch. While most point-releases occur on a four-to-six week schedule, this one took about nine weeks.
The headlining feature from the press release is Ray Traced Distance Field Soft Shadows. In the real world, lights have an actual size. A light bulb is not an infinitesimal point, it fills up your hand when you grab it (when it is off and cooled to roughly room temperature, of course). If a surface can see a light, it is lit by it. If the surface cannot see the light, it is not lit by it, which looks like it is covered in shadow. If a light is big enough that part of it lights a part of a surface, but part of it is blocked, you get "soft shadows".
Ray Traced Distance Field Soft Shadows is a dynamic effect, which means that it can update over time. This is very useful if, for instance, the object that is casting a shadow gets blown up by a rocket launcher or, less entertainingly, the sun sets. The effect is also particularly quick, with scenes rendering in just a couple of milliseconds (you get about 16ms to hit 60 FPS). This is faster than cascaded shadow maps (a method to generate shadows that is optimized for shadows near the camera) in benchmarks listed at Epic's documentation.
Unreal Engine 4.5 has also updated Subsurface Scattering. I am not exactly sure what is different, because Unreal Engine 4 had SSS for quite some time now, but they changed something. This technique is useful to create realistic skin, but is also very useful for oceans, ice, and wax.
Although Ray Traced Distance Field Soft Shadows and Screen-Space Subsurface Scattering are the most interesting feature to write about, I would consider C++ Hot Reload to be the most important feature of this version. To explain it, I will need to first describe how Unreal Engine 4 is designed. When you subscribe, you are given source code access to the engine on GitHub; alternatively, you can download the Unreal Engine Launcher, which allows you to manage canonical builds of Unreal Engine. When a version of the engine is run, it will open a project in Unreal Editor. These projects could be programmed either in C++ or Epic's flowchart-based scripting system, "Blueprints". Complete games could be made in Blueprints, and developers are encouraged to do so, but they are often used for simple objects (lights and elevators), modifications of complex objects, and rapid prototyping.
Rapid prototyping is the key part of my explanation. Remember how there is "engine code" that, when compiled, opens an editor to run "game code" for any given project? Despite the E3 2012 demo, many changes in a project's C++ source require the editor to be shut down and reloaded when game code is compiled. This led people to use Blueprints as a prototyping tool, not because of its logical, visual layout, but because you could manipulate objects several times in just a couple of minutes and without closing the editor. Now C++ is said to be a first-class citizen in this regard (unfortunately I have not had time to test this). As long as you are not modifying the engine's code, just the C++ code associated with your project, your changes should be possible while remaining in editor.
Also updated, and finally supported by default, is Unreal Motion Graphics (UMG). UMG is a UI platform that is built upon Slate, which itself is the main UI platform for Unreal Engine 4 (Unreal Editor, for instance, is created with Slate). Basically, it extends Slate and includes a Flash Professional-like editor for it, complete with styles, animations, and scaling for high-DPI devices.
Because I am not in the DirectX 12 private beta, I am unsure whether that branch has been updated. Microsoft has announced that it was based on Unreal Engine 4.4. They have not said anything publicly since, at least not regarding that.
Unreal Engine 4.5 is available now for subscribers through GitHub or the Unreal Engine Launcher.
- 2 of 2