The Libertarian

Just chatting with a buddy about the prices of DVD players in the US versus here in Denmark. It seems (in Michigan at least) you can get a Philips DVD player with a standard 1 year of warranty for 18$. That seemed extremely cheap compared to here, but for certainty I went ahead and dug up prices on local offers in my area (been a few years since i bought either VHS or DVD players, as the few movies I do watch is on my computer, and I just don’t watch TV so I obviously don’t record anything).

The cheapest DVD player I could find is a DANTAX (ironic name, btw) which is a cheap brand and that was 102$ are current exchange rates. Mainstream brand names are a good deal above that - like LG (134$) or Philips (227$). True, we Danes do get 2 years warranty (a default required by law), but then I asked what the price of an additional 2 year warranty on top of the Philips purchased in the US was. Answer: 8$.

So the approximate difference in consumer price for a bare-bones DVD player looks like a factor of 4.

Damn, we are getting screwed here in Scandinavia…

VW (VolksWagen)’s chairman of the board, Ferdinand Piëch, reveals so in an interview with him on his 70th birthday. He promises that VW will have 100 kpl (km/l) cars on the road in 4 years - by 2011.

Piëch claims that such gasoline-powered, automotive vehicles are now marketable, because the components necessary (he doesn’t say which ones), which costs 35.000 € today, will be obtainable for 5.000 in two years.

Piëch has been calling the shots @ VW after he came on the board instead of Bernd Pischetsrieder, who was the person responsible for killing the Lupo 3L car, which, as the name implies, can drive 100 km on 3 liters of diesel, achieving a 33 km/l fuel efficiency.

Portugal recently opened an 11 MegaWatt solar power center, producing enough power to supply 8000 households. The center consists of 52,000 PV cells covering an area of 0.6 square kilometers, supplied by 4 different companies, it will reduce carbon emissions by 13,000 tons per year as long as it is operational, and produce an expected 20 GWh/yr.

Earlier this year, the EU declared that by 2020, 20% of the EU power consumption would be supplied by renewable energy sources - at the moment, 6.5% of the energy use in Europe is delivered by renewable sources.

Source.

My comment: This is a sign of good progress, but collecting all the PV in one spot is contrary to one of the definite advantages of the technology, which is de-centralization.

IBM is on a rampage!

Now they have introduced vacuum as an insulator between the conductors inside CPUs; the benefits are that processors may now perform up to 35% better, while using 15% less power. Counter-intuitively however, the technology is called “AirGap”.

The manufacturing technique for inserting the vacuum gaps was derived from research on self-assembling molecules, called diblock copolymers, conducted by IBM researcher Chuck Black. A diblock copolymer consists of two types of molecules that, in ordinary circumstances, would repel each other. By designing the molecules in a particular way and controlling how they interact, they create intricate patterns through chemical repulsion.

Since IBM is in a partnership with AMD, they will be allowed to make use of the technology, but not Intel (unless the two parties can come to an agreement, which I hope).

We will see AirGap in IBM’s 32 nm processors which rolls out in 2009.

Also, last month, IBM pioneered a new way of designing “3D” processors, where the chip essentially becomes three-dimensional, and not just a two-dimensional slab of transistors. That means that much more space available for components and thus much more potential processing capacity on a die:

(12th of April) IBM today announced a breakthrough chip-stacking technology in a manufacturing environment that paves the way for three-dimensional chips that will extend Moore’s Law beyond its expected limits. The technology – called “through-silicon vias” — allows different chip components to be packaged much closer together for faster, smaller, and lower-power systems.

…

The new IBM method eliminates the need for long-metal wires that connect today’s 2-D chips together, instead relying on through-silicon vias, which are essentially vertical connections etched through the silicon wafer and filled with metal. These vias allow multiple chips to be stacked together, allowing greater amounts of information to be passed between the chips.

The technique shortens the distance information on a chip needs to travel by 1000 times, and allows for the addition of up to 100 times more channels, or pathways, for that information to flow compared to 2-D chips.

IBM is already running chips using the through-silicon via technology in its manufacturing line and will begin making sample chips using this method available to customers in the second half of 2007, with production in 2008.

This means that multi-core CPUs will be MUCH easier to make, because each core can simply be stacked on top of other cores, reducing chip area.

Science fiction?

Not any more. Using TMS (Transcranial Magnetic Stimulation) humans can now be induced to go into a slow-wave sleep which gives much more of the needed deep sleep, which is essential for our functioning during the day, and memory formation.

At first, it will be of use for insomniacs, but in the future, every home should have one of these devices, so we can enjoy more of life without being comatose.

And a good thing too, because as I wrote a few days ago, they offer more light for the Watt. What holds LED lighting back at the moment is the high price tags they carry, but that will, unsurprisingly, change before too long.

Conventional incandescent/filament lightbulbs are still the most common ones, and they are atrociously inefficient - only 5% of the energy released from an incandescent bulb is in the form of visible light - the remaining 95% is in the form of IR photons - heat. And as if that isn’t bad enough, the old bulbs only have a typical lifetime of 1000 hours.

Enter LED lightbulbs; they have a typical lifetime of 100,000 hours - 100 times longer than incandescent lightbulbs. That, and it offers at least 6 times more visible light per consumed energy compared to incandescent lighting. At the moment, an LED bulb costs around 60 US$, but according to news.com, the price has been halving itself every year, so by 2009, we should be able to get LED bulbs for 15-20$. At that price, the LED bulb pays itself off in saved energy bills after one year, if used heavily (such as in retail stores).

The US DOE (Department of Energy) estimates that 22% of electricity consumption in the US is due to lighting.

If 25 percent of the lightbulbs in the U.S. were converted to LEDs putting out 150 lumens per watt (higher than the commercial standard now), the U.S. as a whole could save $115 billion in utility costs, cumulatively, by 2025, said DenBaars, and it would alleviate the need to build 133 new coal-burning power stations.

This is a typical win-win - good for my wallet and good for the environment. What are we waiting for?

Oh, and another thing - the usual way to make white light by way of LEDs are to combine three LEDs: red, green and blue. This offers interesting opportunities for those of us who like other colors than plain white:

…the Sun!

The (approximate) vital stats of this huge gravitation-powered fusion reactor:

• Diameter: 1.392e6 kilometers (109 Earth diameters)
• Volume: 1.412e15 cubic kilometers
• Surface area: 6.087e9 square kilometers
• Mass: 2e30 kilograms
• Surface temperature: 6000 Kelvin
• Core temperature: 13.6e6 Kelvin
• Instantaneous radiative power: 382.7e24 Watts (make a note of this!)
• Instantaneous mass loss due to mass-energy conversion: 4.3e9 kilograms/sec.

So, what does this mean for humanity?

A damn lot of exploitable energy, that’s what it means!

Outside the atmosphere, the Earth receives circa 1366 Watts/m² (also known as the solar constant), which begins to sound really impressive when you consider that every square meter of Earth’s cross sectional area (= 128e6 km²) receives this amount of energy, for a grand total of 174.08e15 Watts - every second, all year long.

Of course, our atmosphere sucks up a bit of the solar radiation (including the potentially harmful UV-light in the shortwave bands), so we do not see all of this energy on the surface of the Earth. The insolation of Earths surface is slightly less than the solar constant; on a clear day, approximately 16% of the visual-band solar energy is absorbed by the gases and vapors in the atmosphere. That leaves us with 1147.5 Watt/m² to play with, at best.

For those of us who reside in the temperate regions of this Earth, we have to factor in that the sun is never at the zenith of the sky, so we have to angle up our solar receivers to catch as much of the photonic energy as possible. For example, here in Denmark (which is on the 56th northern latitude), a surface that is arranged horizontally on the ground soaks up around 1000 kWh/m²*yr, and if angled up at 45°, it will receive ca. 1200 kWh/m²*yr. Assuming a 12-hour day on average (the definition of the equinox), that means my potential solar panel will, averaging over a whole year, receive ca. 833 or 1000 Watts per square meter, for the above cases. (Again, this assumes that the weather is always clear, which is clearly unreasonable.) A dependable figure for long-term time-averaged insolation of areas typically clear of clouds and haze is 250 Watts/m², which also accounts for the nighttime absence of the suns rays.

In the future, “smart” solar receivers will angle themselves to maximize energy capture, depending on the latitude the solar panel is operating on. In wintertime on the Tropic of Cancer or the Tropic of Capricorn, you need to angle the panel at most 64° from the ground; here in Denmark we would need to angle the receivers up to 79° from the ground for optimal reception. On the other hand, if you live exactly on the equator, you would never need to angle the solar panel more than approx. 23° from the ground for optimal reception! (All this is due to the fact that the Earth’s axis of rotation is off by 23.5° compared to the Ecliptic. Geometry is interesting stuff.. :)

Now, how much energy does that leave us, realistically?

A typical, mass-produced, Poly-Chrystalline Silicon Photo-Voltaic (PV) solar panel today has a practical conversion efficiency that lies between 13 and 16% of the incoming photonic radiation. Of course, there are more advanced types of PV technology out there, which have higher efficiencies, but they are much more expensive and not commonplace, so for now, we shall assume that we can harvest 1/6th of the solar energy we receive from the sky.

250 Watts/m² divided by 6 is 41.67 Watts/m² (as a daily average). Let us assume that humanity as a whole currently consumes 12e12 Watts; that means we would have to produce a lot of PV panels to harvest enough energy for our human purposes:

12e12 Watts / 41.67 Watts/m² = 287.98e9 m² = 287.98e3 km²

So, with current-day technology, we need to erect ca. 288 thousand square kilometers of PV panels in the sunnier spots of the Earth to supply humanity with energy solely from the Sun. More, if we are to keep electricity generation local, to avoid using the lossy power grids.

Let us look further, and try to account for the future:

First, we must assume that humanity’s energy consumption as a whole will increase substantially; not because we in the industrialized world will consume more energy (we are already working on reducing our energy consumption), but because the developing world is rising in living standards, and they will want the same level of comforts as we do today in the western world; cars, computers, lighting, air conditioning and so on. And also, because the human population of Earth will probably double from today’s 6 billion to around 12 billion by 2100 or so. All these people need energy to go about their daily lives.

If we choose a round number and assume that to total energy use of humans will increase by a factor of 10x the next 50 years, then we will need 2.88 million square kilometers of PV panels to supply the Earth with power. This sounds like an awful lot, but keep in mind that the total surface area of the Earth is 2.82e17 square kilometers, so if we are to succeed in this mission, we would actually only cover a hundredth of a billionth of the total surface area of the Earth. Or expressed in figures: ≈ 0.00000000001 or 0.000000001%.

Now, that was an estimate in the pessimistic direction - lets look the other way:

PV technology will surely improve!

Already we have PV tech with an efficiency rating of around 33%, and with the nanotechnological fabrication advances that are constantly being made, we are closing on the magical 50% mark. Assuming we reach that, we would only need a third of the above total PV panel area to satisfy the total future energy needs of humanity - and that would amount to 960,000 square kilometers.

If we shoot high and hope for a very-high efficiency PV technology to be available to us in the future, 83% efficiencies are not reserved for the realm of the fantastic - with that, we would “only” need 576,000 square kilometers of PV paneling spread around the Earth.

With the future offering us unimaginably colossal manufacturing capacity via fully automated factories and molecular assemblers, this shouldn’t be an unreasonable goal for humanity. All we lack is the will to proceed.

Perhaps the word “tech” is overstating it, but anyway; IBM discovered that if the surface of the chip that needs cooling is etched with very fine grooves, the thermal paste between the heat source and the heat sink conducts heat better, and thus leads to better cooling:

(…) Big Blue boffins at the company’s Zurich Research Lab discovered that if the upper surface of a chip’s ceramic casing is etched with a network of channels, the thermal paste applied to conductively connect the chip to its cooling mechanism - heatsink, fan, whatever - spreads in a more efficient manner. Or, as we say in the trade, it gets thinner.

The upshot: the metal particles suspended within the paste are better able to conduct heat away from the chip.

This means we will soon be able to pump a higher wattage into the CPU, and thus get more performance out of it without it overheating and shutting down in a particularly nasty system crash. Or, for those of us who aren’t die-hard CPU over-clockers and computing speed freaks, we can get the same amount of chip cooling with lower heat sink fan speeds (meaning a lower system noise level) or… even go all the way to passive, noiseless cooling.

Thank you, IBM.

Not new, but important; in June, 2006, Cree Inc. created a very-high energy-efficiency LED which now makes domestic lighting much more economical, both for our wallets and the environment that is affected by the coal-burning power plants that produce the energy that your household consumes:

MANHASSET, N.Y. — Cree Inc. said it has produced a white LED with efficiency of 131 lumens per watt, confirmed by the National Institute of Standards and Technology.

“This is the highest level of efficacy that has been publicly reported for a white LED and raises the bar for the LED industry,” said Scott Schwab, Cree general manager, LED chips, in a statement.

Semiconductor suppliers have racing to produce higher efficiency white LEDs as the industry seeks energy-efficient alternatives to conventional lighting. In March, Japan-based Nichia Corp. reported it had developed a white LED rated 100 lumens per watt.

Last September, Cree (Durham, N.C.) said its white XLamp 7090 Power LED was capable of producing 86 lumens per watt.

Lumens-per-watt is the standard used by the lighting industry to measure the conversion of electrical energy to light. As a reference, conventional incandescent light bulbs are typically in the 10 to 20 lumens per watt range, while compact fluorescent lamps range from 50 to 60 lumens per watt.

IOW - using LED lighting will get you between 6 and 13 times more light per watt than old-fashioned incandescent bulbs and up to twice as much as existing energy-saving fluorescent tubes.

The chip above measures 3.25 by 5.25 millimeters and promises 160 Gbps data transfers, perhaps as early as 2010. Keep in mind that 160 Gigabits are 20 GigaBytes (approximately the same a one H.264-compressed HD-Video full-length motion picture) - and this miniature chipset would allow that amount of data to be shunted through to your computer every second.

Excuse me while I go into spasms of geeky joy.

There is one problem, though; and that is that singular hard drive mechanisms today are much too slow to make use of even one percent of this potential transmission speed. The best 7200 rpm hard drives can read and write approximately 60 MegaBytes of data per second, and the fastest consumer storage bus, SATA-II, can only pump 3 Gigabits though per second, and even that we are far away from fully exploiting today.

More.

InPhase shows off a 1.5 millimeter-trick, 300 GiB capable holographic storage disk, which we are promised will scale up to 1.6 TB by 2010. Data transfer rates from the disk are only 20 MiBps at the moment, though. A ReWriteable version expected in 2008. Shelf life: 50 years.
Price: 18000$ per drive, 180$ per disk.

Mempile presents something somewhat better, a 500 GiB holographic disk sporting 100 virtual layers each of 5 GiB capacity. Future versions will allow 200 layers… truly making it a TeraDisc.

Note: What makes these disks holographic is that they are transparent - there is no reflective layer in the disks, like in CD’s and DVD’s.

Who needs BluRay?

Inertial Electrostatic Fusion systems can now be built

http://www.fusor.net/board/view.php?site=fusor&bn=fusor_announce&key=1143684406

Our company, EMC2, has been working since 1987 on the R&D of Iour polyhedral IEF concept for fusion; mostly under DoD support.

Final tests were made last Oct/Nov on a unique new design, based on unexpected discoveries made in Spring/Summer 2005.

This final machine, WB-6, showed 10x lower e- losses than any predecessor and produced DD fusions at a rate over 100,000x times higher than the data of Farnsworth-Hirsch in the 1960’s for same drive conditions.

We have now proven the engineering and physics scaling laws that allow design of full-scale net-power systems, whether on DD or pB11. USNavy budget line item that supported our work was zero-funded in FY2006, and our lab had to shut down and close one week after achieving these results!

We are probably the only people on the planet who know how to make a real net power clean fusion system, and we are out of support! Somewhat ironical!

The next logical step MUST be a full-scale net-power demo system, simply because there is not much left to do at small scale; when it is realized that the fusion output of these devices scales as the 7th power of the size, and the gain scales as the 5th power.

These outlandish scalings (inherent in the engineering physics of the thing) make it useless to build half-scale systems (for example). Unless you are AT the net-power size, you are nowhere in power and gain, even though the physics IS relevant. We have always been limited to about 0.1 scale, and have learned nearly all there is to know about the system’s basic operation.

Thus, we have the ability to do away with oil (and other fossil fuels) but it will take 4-6 years and ca. 100-200 M$ to build the full-scale plant and demonstrate it. Anyone care?

- R.W. Bussard

(posted in sci.econ)

http://en.wikipedia.org/wiki/Robert_W._Bussard
http://en.wikipedia.org/wiki/Farnsworth-Hirsch_Fusor

Of course, we need to do away with the Arabian-American oil interests before this comes to the market in commercial quantities… :/

Unknown to some (myself included, until last week :), there is a copper-based LAN technology out there that is faster than the current “top-of-the-line” 1 Gbps Ethernet: It’s 10 Gbps Ethernet, potentially 10 times faster than common consumer Gigabit Ethernet, and it is extremely expensive; one specimen of a 10 Gbps LAN adapter is Intel’s PRO 10 GbE CX4 Server Adapter, which costs to the tune of 1000 US$. 10 Gbps switches are also outrageously pricey - the cheapest switch I have been able to find is SMC’s 8-port TigerSwitch SMC8708L2, which sells for around 6000 US$!

But that’s not all - this year, there have been various efforts to bump up the speed of Ethernet further; IEEE recently announced that a study group to design the next generation of Ethernet has now agreed on the speed of this technology, which has been nailed down to 100 Gbps (!!!). The standard should be finalized by 2010.

Before you start choking on your own drool in your expectation to see 10 or 100 Gbps Ethernet connectivity in your home in a few years time, I have to disappoint you; if the price was not hint enough, this kind of equipment is practically reserved for ISP’s and really big businesses; essentially, companies that have extreme network connectivity requirements. In the foreseeable future, consumers and even “prosumers” such as you and I, will realistically have no need for the amount of bandwidth that these technologies offer.

Just look at the hard drive in your computer: Most hard drives’ data transfer speeds max out at around 50 MBps (= 400 Mbps) these days (due to drive mechanism limitations), which is around half of the capacity of Gigabit Ethernet. Even when everyone has 10.000 rpm RaptorX (which can deliver around 70 MBps = 560 Mbps) hard drives installed in their computers, Gigabit Ethernet will STILL be enough for all but the most demanding users!

Besides, when it’s time to move beyond (1) Gigabit Ethernet, I’m confident that we will also be going beyond copper and electrons to carry data from point A to point B - we will be adopting optical fiber networking, even in the home of regular Joes.

But the aforementioned ISP’s and big businesses that need extreme bandwidth will have use for not only the 10 Gbps Ethernet of today, but also the 100 Gbps Ethernet of tomorrow, seeing how consumer broadband is taking hold around the world, and high-speed fiber connections are coming to market these years. All that is needed to saturate a 10 Gbps interface is approximately one hundred 10 Mbps clients, and there are already a god deal of these, thanks to the recent advances in DSL and cable technology.

Lastly, while PCI-X@133 MHz is barely enough to saturate a 10 Gbps link, and the top-spec 16x PCIe is nowhere near enough to saturate a 100 Gbps link, we will also have to see enhancements in the area of system interconnects before we will be able to deploy the latter technology, but there shouldn’t be any doubt that that will happen, as well.

The future of transportation is clean, and Honda is in the lead of the pack of companies that make use of the new hydrogen fuel-cell technology.

Honda’s FCX prototype is looking like a worthy contender for the future of the automotive industry. The 2005 model of this ZEV (Zero-Emission Vehicle) is a roomy, streamlined beauty of a car, with these specs:

• 80 kW / 107 hp AC electric engine
• 272 Nm torque
• 156.6 liters of hydrogen fuel stored at 5000 psi
• 190 mile operating range

Sure, that’s not much compared to current-day gasoline-combustion cars, however keep in mind that this is only the second generation of fuel-cell technology for automobiles, and we will see more improvements in the next ten years, when these technological gems are expected to make their entry into the marketplace. The FCX/2005 improved over the 2004 model by increasing fuel economy (and thus, operating range) by 20%.

I can’t wait to try one of these engineering masterpieces on the road.