August 11, 2006

Two-mode hybrid from GM, BMW, DaimlerChrysler

DM, BMW and Daimlerchrysler are working and investing together to catch up in hybrid technology. The new two-mode hybrid that they will rollout provides 25% fuel efficiency gain for the big SUVs and cars that they make. Current hybrids are not good for staying efficient and towing or moving heavy loads or vehicles. The goal is clearly to provide more efficiency with the money making big vehicles that people have been buying. If the transmission works as promised, the 25 percent gain in fuel economy would let GM SUVs, such as the 2,720kg Chevrolet Tahoe, get about 26 miles per gallon (9 liters per 100km) in combined city/highway driving. That would likely be enough to keep GM ahead of any US-government-imposed increases in corporate average fuel economy.

Sprint/Nextel announce Wimax network plan for late 2007

Sprint/Nextel will rollout a Wimax network with peak communication speed of 20 mbps but with real world speed of 1-4 mbps. They expect to spend $3 billion and go live late in 2007. The company will use its existing 2.5GHz spectrum, half of which it acquired from the merger with Nextel, to deliver the new service. The service will be three to ten times faster than current generation EV-DO. They and supplying companies (like Intel, motorola, Samsung) claim the chipsets that will be used in the devices and to build the infrastructure equipment will cost about a 10th the price of existing 3G chipsets. Sprint plans to upgrade its EV-DO system to EV-DO-A.

Proposed Solution for Cumbre Vieja

Thinking about the prior article about nuking Cumbre Vieja and other targets.

I think the solution for risky natural features is like what we do with avalanche management.

You need to identify such risks and start removing them in a controlled way.
Blow off smaller chunks (or massive excavation) that would not cause a disaster.
Obviously be careful doing it.

Japanese researchers make adult cells behave like embryonic stem cells

Adult cells have been transformed to take on the character of embryonic stem cells. The cells--which the researchers designate "induced pluripotent stem cells" (iPS)--exhibit the physical, growth, and genetic characteristics typical of embryonic stem cells, they reported. "Pluripotent" refers to the ability to differentiate into most other cell types.

nanoscale: Modeling software could lead to more efficient design of nanoparticles.

Accelrys, a San Diego-based company with experience developing modeling software used for designing drugs and materials, is building modeling software to help researchers develop new materials and test their compatibility with various molecules, potentially speeding up the development of nanotechnology for drug delivery and diagnostics.

The software would help researchers sort through many combinations of drugs and polymers, to predict which ones will be compatible with each other and which can be made into nanoparticles that safely deliver the drug to target tissues, such a cancer tumor. They can match materials with protein-detecting molecules to help in designing diagnostic nanoparticles that latch onto molecular biomarkers in the body and reveal their location with various imaging technologies.

MIT Energy Manhattan projects

MIT energy manhattan project described at Wired Ideas include lithium nickel manganese oxide batteries for better hybrid cars. Plasmatron boost to cars providing near Hybrid level mileage for only $500-1000 per car.

August 10, 2006

Automatic Hardware acceleration methods 10 to 30 times faster

A discussion of encoding with automated methods (so no hand optimization) software algorithms in fpga hardware and getting 10 to 30 times faster. Recent advances in C-to-FPGA design methodologies and tools facilitate the rapid creation of hardware-accelerated embedded systems.

C to HDL design flow Converting C code to an HDL (Hardware Description Language) accelerator with a C-to-HDL tool is an efficient method for creating hardware coprocessors.

This is relevant for supercomputers and getting flexible and cheap speed for commonly used software.

August 09, 2006

Cobalt Green allows spintronics to work at room temperature

18th century paint Cobalt Green will allow spintronics to work at room temperature To test the antique pigment the researchers processed zinc oxide, a simple semiconductor material, so that some of the zinc ions were replaced with ions of magnetic cobalt. The process is known as "doping". The cobalt ions were then aligned by exposing the semiconductor to a zinc metal vapour. The alignment causes the material to become magnetic. The magnetism continued when the material was warmed to room temperature and when the exposure to the zinc metal vapour was stopped.

The research is still in its early stages. To be of use to chip-makers the technique must be shown to work and integrate with silicon semiconductors.

Gene revolution could help the poor

UN says gene revolution could help feed the poor with more productive crops I agree.

Drought and insect-resistant crops could boost yields and incomes.

Critics talk about the problems of distribution of food and lack of political will. Sure, but better crops will help. China does not have big distribution problems but has had famines. Better crops will reduce famine there. Better crops and more excess makes for a system that is more resistant to distribution and corruption problems. Not immune but more resistant.

Ovonic Cognitive computer perhaps what the Ovonic quantum control makes possible?

Here is a pdf from 2004 where Ovshinksy discusses the Ovonic Cognitive computer A single Ovonic device (or in some cases, two devices) of subnanometer size is able to have many multiple functions such as the demonstration of addition, subtraction, multiplication and division, along with the standard binary activity of any computer, it also can do nonbinary processing, modular arithmetic and encryption as well as factoring. This goes along with being potentially cheaper and low power usage and enabling an all thin film computer.

The device has neuron like features. It can operate at smaller than 10 nanometers and can be used by itself or can be mixed in with semiconductors. The neuron like features could enable advances towards far better artificial intelligence.

IV curves of Multi-terminal OQCD (Ovonics quantum control device)

Previous articles:
Article 2 on the ovonic quantum control

First article on the announcement

More on Ovonic Quantum Control, potential transistor replacement

It sounds like it could be cheaper and more energy efficient than transistors.

More on Ovonic quantum control, potential transistor replacement. The all thin-film device is said to have significant multifunctional capabilities when compared to transistors, thanks to its high current carrying capacity and unique modulation gain. Its multifunctional operating modes include the ability to be turned on using a small pulse applied to a third terminal in either latching or non-latching manners. Using thin film fabrication, Ovshinsky says the device can be fabricated in a cost-effective manner, using the Ovonic roll-to-roll process, a continuous web, triple-injection, roll-to-roll photovoltaic processor already makes 9 miles of thin-film, semi conducting photovoltaics in a single run.

Picture of the inventor, Stanford R. Ovshinsky, ECD Ovonics president, chief scientist and technologist, and a wafer with ovonic quantum control devices on it

Related article:
Past article on the announcement of the device a couple of months ago

An article about the possible Flash replacement memory created by the same inventor and the comparison of that technology to other computer memory

Nanotubes and semiconductor memories and other tech could be coming soon

Researchers at the University of Cambridge have successfully grown nanotubes (M. Cantoro et al. "Catalytic chemical vapor deposition of single-wall carbon nanotubes at low temperatures", Nano Letters 6, 1107-1112, 2006) at the lower temperatures needed for full integration into present complementary metal-oxide semiconductor (CMOS) technology (350ÂșC).

Nantero talked about releasing its hybrid nanotube semiconductor memory.

other tech: fairly practical flying car finally

Fabric wing flying cars are here. New lighter fuel efficient cars would be modifiable to flying easily. If the FAA, insurance companies and other regulatory bodies would allow it. It looks like it would be a somewhat practicle flying car where the car temporarily converts.

Atair is also working on a two-person paragliding buggy called Chimera, which has a propeller at the back and powered wheels.

More on the tactical capabilities Adding a small 30 HP engine with propeller to the HAHO Pod adds additional strategic capabilities. The engine provides propulsion for extended and ascending flight. The system can take off again once landed, being airborne in 50-150 feet; a paved runway is not required.

It could be parachuted from an aircraft and travel around on the ground, becoming airborne when necessary by redeploying its canopy. It will travel at about 46.3 kilometres per hour and consume 7.5 litres of fuel per hour, the company says.

other tech; Ballbot

Carnegie Mellon University researchers have developed a new type of mobile robot that balances on a ball instead of legs or wheels. "Ballbot" is a self-contained, battery-operated, omnidirectional robot that balances dynamically on a single urethane-coated metal sphere. It weighs 95 pounds and is the approximate height and width of a person. Because of its long, thin shape and ability to maneuver in tight spaces, it has the potential to function better than current robots can in environments with people.

Nanowire 'barcode' system speeds biodetection in the field

Detecting biowarfare agents in the field will become a lot easier thanks to a new barcode system based on biosensing nanowires developed by Lawrence Livermore National Laboratory researchers. Multi-striped nanowires developed at LLNL allow rapid and sensitive immunoassays for biowarfare agent simulants.

The researchers produced nanoscale wires by electrochemically depositing metals within the tiny cavities of porous mineral solids. They then layered the gold and silver in a specific way to produce nanowires with different characteristic stripe patterns depending on which pathogen they were trying to identify.

The reflection pattern and fluorescence from each stripe sequence can later be clearly recognized, similar to a barcode on a retail product.

“Antibodies of specific pathogens have been attached to the wires,” said Jeffrey Tok, principal author from LLNL. “This produces a small, reliable, sensitive detection system that can easily be taken into the field.”

The system not only applies to biowarfare agents, but could also be used during an outbreak of an infectious disease.

Hybrid semiconductor molecular chip designed

A proposal has been written for a reviewed journal for a hybrid semiconductor and molecular chip that would have 100 billion bits per square centimeter of flash-like memory.

In this cross-section of a portion of the hypothesized hybrid micro-nano-molecular IC, a crossbar (vertical lines) is placed on top of conventional CMOS circuitry. The inset drawing shows the molecules that link the lower and upper sides of the crossbar and “mimic the behavior of a flash memory cell.”

Production cost and scalability are the key challenges. Other questions – such as reliability of the molecules –would require further specifications.

August 08, 2006

Diamond technology to revolutionize mobile communications

A new technology based on Ultrananocrystalline DiamondTM (UNCDTM), a novel material developed at Argonne that will enable diamond resonators and oscillators to be directly integrated with microelectronics chips for next-generation telecommunication devices. These devices will be fully integrated with silicon microchips to enable a new generation of high performance portable communication systems (maybe 20 or more times faster than current technology). Eventually the fruits of this project could result in enabling a variety of mobile technologies with much higher data communication rates. This hybrid approach seems likely to help tap the benefits of diamond materials in more situations and at lower costs.

Related article:

Diamond semiconductors.

August 07, 2006

Only have one fairly big nuke, where would it do the most damage?

A list of targets that have multiplier effects of follow on devastation.

I am not convinced about the Iran scenario.
Other targets with multipliers. Tokyo, Saudi Oil Fields.

End of the world threats rated

Each threat is estimated in two ways: first, the chance of it occurring in our lifetime (the next 70 years); and, second, the danger that it would pose to the human race if it did happen (10 = making humans extinct, to one = barely having an impact on our lives).

Threat Likelihood in 70 yrs Impact (10 humans extinct/1 barely any)
1: Climate Change High 6
2: Telomere erosion Low 8
3: Viral Pandemic Very High 3
4: Terrorism Very High 2
5: Nuclear war Low 8
6: Meteorite impact Medium 5
7: Robots taking over High 8
8: Cosmic ray blast from exploding star
Low 4
9: Super-volcanos Very High 7
10: Earth swallowed by a black hole
Exceedingly low 10

Other threats:
Nanotech war and/or Space war using kinetic bombardment
Chance (no chance before better tech in 15-35 years/ after depends on politics and people)
Impact 7-10

TEM microscope advance 1.4 Angstrom resolution at less damaging (to sample) voltage

For the first time ever, directly interpretable (Transmission Electron Microscope) TEM images with atomic resolution better than 1.4 Angstrom were obtained at the very low operating voltage of 80kV.

The result was welcomed by some of the world's leading research centers as an important milestone in nanocharacterization as now even light element materials such as carbon nanotubes and graphene can be imaged artifact-free and with high contrast while having highest lateral resolution.

Direct atomic resolution at 80kV was obtained for various classes of materials: gold nanoparticles, silicon and single wall carbon nanotubes. The smallest atomic distance resolved was the well-known silicon dumbbell distance of 1.36 Angstrom.

Combining protein fragments to tailor/customize mechanical properties

Combining proteins can be used to tailor mechanical properties of the resulting protein. A team from the University of British Columbia (Vancouver, Canada) succeeded in producing proteins with new mechanical properties through the combination of two "parent" protein fragments.

The research group of Hongbin Li chose two different titin domains from heart muscle for their experiments. Titin, a giant molecule, is responsible for controlling the passive tension of our muscles and also pulls them together again after an extension.

As the parent generation, the scientists chose two globular titin domains called I27 and I32, whose mechanical properties have already been intensively researched. Both are similarly built and are composed of the protein segments A, A’, as well as B to G. The researchers interchanged several fragments of the genes that encode I27 and I32 ("DNA shuffling"). Genetically they produced four different protein "children": an I27 with the A’/G strands from I32, an I32 with A’/G from I27, an I27 with C, D, and E from I32, and also an I32 with the C, D, and E strands from I27.

The mechanical properties of all the proteins were investigated with atomic force microscopy. To do this, one end of the protein chain was attached onto a solid support and the other end was adsorbed onto the tip of the atomic force microscope. When the tip is gradually pulled away from the support, the protein elongates and the force increases until the protein finally unfolds. The resultant force–extension curves characterize the mechanical properties of the proteins. It turned out that all children show different mechanical characteristics to their parents. It was previously thought that the specific arrangement of the A’/G section was critical for the mechanical stability of the domain, whereas other parts of the domain, including the C, D, and E strands, only played a less- significant role. This opinion must now be revised.

Li hopes that this exciting new application of the powerful recombination technique in the field of protein mechanics will open a new way to tailor proteins’ mechanical properties.

New Petaflop computer planned for 2008

The Energy Department’s Oak Ridge National Laboratory and Cray announced a $200 million deal in June to make a petaflop computer in 2008. The computer code, dubbed Qbox, will help researchers run science simulations deemed essential to national security. The supercomputer, which Cray nicknamed Baker, will use optimized Advanced Micro Devices dual-core Opteron processors.

The current fastest computer is the japanese MD Grape 3 which is already at a petaflop.
The fastest computer that can run standard benchmark code is the Blue Gene computer.

DNA overwinds until over 30 piconewtons of strecthing is applied

DNA overwinds when stretched instead of unwinding. It does not unwind until more force is applied. Overwinding DNA could be used to power nanomotors.

To test what happens when a DNA molecule is overwound, the DNA was stretched between a glass coverslip and a paramagnetic bead, while a fluorescent avidin-coated rotor bead was attached to the DNA just below a biochemical nick. Tension in the DNA was controlled by raising or lowering the magnets, and changes in winding were observed by tracking the rotation of the fluorescent bead.

A simple toy model shows DNA as an elastic rod (grey) wrapped helically by a stiff wire (red). Stretching generates an overwinding of the helix because the inner rod decreases in diameter as it is stretched. The outer helix is then able to wrap a larger number of times over the length of the molecule

To explain the overwinding, Bustamante and his coauthors have proposed a simple “toy” model in which the radius of the DNA double-helix is allowed to shrink as the molecule is stretched. The model consists of an elastic rod that is wrapped around its outer surface by a stiff wire, analogous to DNA’s sugar-phosphate backbone. The elastic rod is constructed from a material that conserves volume under stress.

“As this system is stretched, the elastic rod decreases in diameter,” said Bustamante. “This enables the outer wire to wrap a larger number of times over the length of the rod.”

The twist-stretch coupling results demonstrated by Bustamante and his collaborators holds important implications for how DNA-binding proteins are able to recognize their target sites along the helix. These proteins are known to bend, wrap, loop and twist DNA. Now it has been shown that they can achieve their goals by simultaneously stretching and overwinding a DNA molecule, or by compressing and underwinding it.

August 06, 2006

Transhuman: Iron Man versus Borg versus created X-men

Many groups are concerned about transhumanism and nanotechnology. They are worried for various reasons about people being made stronger, smarter and with more capability than others. They fear the creation of supermen who will dominate others or that the act of creating superhuman abilities would be a transgression against religion.

I think it would be helpful to analyze what capabilities make sense to put into the body like Star Treks Borg, (cyborgs like Robocop or Steve Austin, the 6 million dollar man), put into the cells and genetic code (like artificial X-men) and what makes sense to leave as a wearable tools (like the comic book character Iron Man).

We can also analyze what enhancements might make a difference or are flashy but mostly irrelevent. One of the main aspects of relevance would be difference in productivity and ability to earn a living of future Iron men versus future x-men versus future Borg. Note: there is no reason for the eventual appearance of any enhancement architecture to be as obvious as the fictional examples.

The analysis can also be stated as: when would it be advantageous to get closer to our technology ? When can we do just fine using technology but not becoming our technology?

One of the things that was mentioned was night vision.

Now: We have night vision goggles now which you can buy on Ebay and other places.

There is limited advantage or difference for this capability. There is less reason to out it in the body versus using the tools.

Strength enhancement.
Mundane Iron Man Now: We have fork lifts and cranes.
Mundane X-men Now: We have steroids, exercise and supplements (creatine, protein powders etc...). In mice they have modified genetics for strength enhancement and I expect that it will work on humans.
More exotic: We are making exoskeletons. (more advanced Iron man and Manga style robotech/Gundam). We will perfect the genetic modifications and the techniques used to deliver those modifications.

Utility: strength helps in various situations. The main reasons for doing it are for general resilience and for an "always on" ease of use. Good exoskeletons would limit the advantage for genetically modifications in this area.

If you can make modifications that are as safe as supplements and with health benefits instead of risks, then why would you not do it? The advantage to be able to lift something is irrelevant because you could get an easily available tool or exoskeleton to help you.

Speed enhancement:
Mundane now: segways, bikes, cars and planes.
Exotic now: exoskeletons.

Utility: Same as for strength. You can do it but the advantages will be limited. There would be an advantage for enhancing reflexes and reaction times. Reaction times would benefit from sensory enhancements. Being able to spot and identify targets sooner.

Intelligence enhancement:
Mundane now: computers. PCs, iPods, there are wearable computers and displays that go directly to the eye.
Exotic now: There is Brain gate and other close interfaces between machine and brain. Many of those are non-invasive.
Utility: The difference between invasive and non-invasive is one of bandwidth and communication speed. There are also advantages to integrated control.
Productivity: This is one enhancement that could have a substantial productivity variance depending upon architecture.
This is one area where getting optimum performance irregardless of architecture will make a difference.

Life extension and regeneration:
Mundane now: drugs, vaccines, surgery, hearing aids, mechanical hearts, prosthetics, enhanced prosthetics, biosensors
Controversial now and soon: stem cells, gene therapy, iRNA
This is another where the genetic and invasive modifications are required.

It would make sense to modify our biology to make us more resistent to radiation, disease and better adapted to space and other conditions.

Uploading/mind transfer:
There are questions as to how well this would work in terms of consciousness. Eventually this architecture could diverge from the cyborg, genetic enhancement capabilities. The communication between biology and the computer and whether upgrading hybrid biology would be slower than pure computer equipment would be factors in whether architectures diverge in performance.

Related Articles and more reading:
Article in Slate talking about transhumanism

Transhuman ethics article from Reason

Transhuman reading list

other: US patent system overhaul coming

The US patent system has significant legislation coming. I think a lot of the first draft will get passed with amendments and fine tuning.

The main elements of the new proposed legislation:
- Change to first file from first to invent.

- 12 month "postgrant opposition" system that would allow outsiders to dispute the validity of a patent before a board of administrative judges within the Patent Office, rather than in the traditional court system. The idea behind such a proceeding, also endorsed by the Patent Office, is to stave off excessive litigation. Can get extension of the opposition but only to limited aspects after 12 months.

- the Hatch-Leahy bill would place new restrictions on the courts where patent cases could be filed. (Restricts "forum shopping" for districts known for favorable judges).

- Curb the amount of damages for winners of infringement suits.

- Perhaps most notably, and in a departure from the House version, courts would have to calculate the royalties owed by infringers based solely on the economic value of the "novel and nonobvious features" covered by the disputed patent, not on the value of the product as a whole.

=====More comments:

First draft, probably because of competing priorities among the technology, drug, biotechnology and other patent-heavy industries they will fine tune. I expect that a lot of what is in it will probably get through. But fine tuned. This will favor big companies and those who execute on deliverying innovation more instead of those who invented innovation first.

The strategy is patent as fast as possible and exploit your innovations and make your money. It will tough for someone to take it away and you will need money to keep others from violating. Only really gross and obvious patent violations will get stopped. You probably have to have a series of patents on the innovations. The minor tweaks will no longer be worth much. (although the tweaks could make a lot of money in business, the legal rewards for the innovation will be a lot less if you have to prove how much came from the tweak)