CPU's shed light

[verndewd]

By verndewd
Dewds news @ xcpus, lead news dawg


A lot of data is beginning to surface from places like IEEE and CIE, about the possible next generation of computing cores referred to as CPU's.

Recent developments have shown that rare earths used in the production of light emission and lasers can be utilized on silicon and carbon nanotubes. A few months back Intel provided diagrams of how an optical transceiver could be used in a multicore chip configuration. One place that set of ideas surfaced was at extreme tech.

Figure 1. Cross-section schematic of a hybrid silicon laser.



In an article I cannot seem to locate there was a graphical representation of how an optical transceiver would fit into an 80 core configuration with GPU units as well. The optical transceiver was aligned as a system bus.
http://www.extremetech.com/art.../0,1558,2093952,00.asp]Intel Showcases 80-Core Research Chip[/url] Feb 12, 2007

http://www.intel.com/technolog.../hybrid-laser-1006.htm]A Hybrid Silicon Laser: Silicon Photonics Technology for Future Tera-Scale Computing[/url]

Figure 2. When voltage is applied to the contacts, current flows, and the
electrons (-) and holes (+) recombine in the center and generate light.

http://www.intel.com/research/...erascale/teraflops.htm]Teraflops Research Chip[/url]

http://www.intel.com/research/...orm/sp/hybridlaser.htm]Hybrid Silicon Laser – Intel Platform Research[/url]

http://img201.imageshack.us/my...jbchip11345300efl3.jpg]

[/URL]


http://www.reed-electronics.co...445429?industryid=3028"]Semiconductor International recently had another article on this subject.[/URL]

nanopillars

quantum dots

Wider bandgap III-V nitrides operating in the visible to deep ultraviolet (DUV) were used to process a range of blue and UV light-emitting diodes and, recently, avalanche photodiodes that may one day offer a more robust alternative to the current photomultiplier (vacuum) tube technology.

a laser lattice

Here we have Quantum name attachments to the ideas. Quantum computers as you know utilize Qbits to process data. And there are cascading laser references, while its not the profound IEEE articles on CPU futures The info isnt directly related, but, it is a late development in laser technology. And I personally feel as if every small or large gathering of related information adds to the whole.

IEEE has an amazing gathering of minds. Back when I and many others were tossing out ideas of optical mother boards, these guys were doing the real work both on paper and in the lab.

From IEEE spectrum (this month's issue oct 05):

http://spectrum.ieee.org/oct05/1915

2005 study on IIIv
http://download.intel.com/technology/silicon/CSICS_2005_paper.pdf

To siliconize photonics, you need six basic building blocks.
An inexpensive light source.

Devices that route, split, and direct light on the silicon chip.

A modulator to encode or modulate data into the optical signal.

A photodetector to convert the optical signal back into electrical bits.

Low-cost, high-volume assembly methods.

Supporting electronics for intelligence and photonics control.

The relentless push of Moore's Law has allowed data rates to soar, Internet traffic to swell, and wired and wireless technology to cover continents. Increasingly, we all expect fast, free-flowing bandwidth whenever and wherever we connect with the world. Within the next decade, the circuitry embodied by a rack of today's servers, able to churn through billions of bits of data per second and handle all the data-processing needs of a small company, will fit neatly on a single silicon chip half the size of a postage stamp.

Over the years, IEEE has published some articles of profoundly awesome Ideas in optical computing; from optical motherboards to transceivers to CPU optics and silicon light emitting chips.

If you have not read them you should. Provided your brain can handle the melting temperatures of the hot ideas.

I have been following the late breaking news on what optical technology means to the future of computing since Intel released their laser transceiver. Shortly before that I postulated the ramifications of it in other threads and imagined it a couple years before that. In part 2 we will go in depth on the IEEE articles regarding this awesome development.

This article is property of Dewds news, and its contents are property of the respective websites listed. Copyright 2007.

Intel Builds New Laser Based Processor
By Scott M. Fulton, III, BetaNews
September 18, 2006, 4:16 PM
UPDATED Researchers with the University of California at Santa Barbara, working in conjunction with Intel, announced Monday the next step in their joint plans to produce an entirely solid-state photonic processor assembly - a chip which processes data as light waves, without the need for microscopic, yet movable, parts.

The last major hurdle to being able to produce a fully fabricated, solid-state optical processor using on-board lasers involved the bonding process, it turned out.

http://spectrum.ieee.org/oct05/1886

Light From Silicon
By Salvatore Coffa
For decades, silicon was a semiconducting dim bulb, but now we can make it into LEDs that match the best made from more exotic materials.

Silicon's absence from critical optical applications has long bothered semiconductor specialists. If photons could be easily coaxed from silicon, we could do marvelous things. Imagine plugging your office PC into an optical-fiber local area network and pulling files from a distant server at tens of gigabits per second?enormous, high-definition video files popping onto the screen instantaneously. Optical fibers linking the microchips within a PC would accelerate its computing speed as bandwidth bottlenecks from its motherboard's copper wiring disappeared.

The key to that vision is the fabrication of efficient, electrically driven light sources that work at room temperature and are produced using materials and processes compatible with the manufacturing methods currently used to make ordinary silicon memory and microprocessor chips.

so here we have the reason for the mad dash, Throughput and possibly computation as well.
Optical signal is derived from rare earth metals that when applied with an elecrtic charge can produce light. As we see in this articles example of Galuim arsenide and indium phosphide.

But lasers and LEDs are made of exotic substances called III-V semiconductors?from their columns on the periodic table of the elements. These materials, which include gallium arsenide and indium phosphide...........

To unite the worlds of microprocessors and lasers, we need cheap, integrated optoelectronics made from silicon........
Intel Corp., in Santa Clara, Calif., for example, announced in January that it had found a way to power a silicon-based laser with a conventional one. The technique allows engineers to integrate the silicon laser on the same chip with such standard computing fare as logic circuits and memory cells, as well as critical optical components such as the modulators that encode electronic bits onto the light beam. But the scheme does not eliminate the rather expensive III-V semiconductor laser; it just makes it cheaper to use. To be rid of the costly III-V compounds altogether, you'd need a silicon chip that turns electricity directly into laser light.

We are almost there. Although the green glowing device my group built in 2001 in our laboratory?which is part of STMicroelectronics NV, the Geneva-based semiconductor giant?did not emit light that was coherent, collimated, and monochromatic (it wasn't a laser, in other words), as a light emitter, it did match the efficiency of conventional LEDs fabricated from III-V semiconductors. Since then, we've been working to make our LEDs more laserlike, and we believe an electrically powered silicon laser?with all that means for computing and communications?is finally within reach.

For a non cut and paste compilation read the article. The next section talks about the specifics of the band gap and valence band traits that make silicon a poor laser and what is bieng done to correct it.

In solid state physics and related applied fields, the band gap, also called an energy gap or stop band, is a region where a particle or quasiparticle is forbidden from propagating. For insulators and semiconductors, the band gap generally refers to the energy difference between the top of the valence band and the bottom of the conduction band.


[edit] In semiconductor physics

Semiconductor band structure.In semiconductors and insulators, electrons are confined to a number of bands of energy, and forbidden from other regions. The term "band gap" refers to the energy difference between the top of the valence band and the bottom of the conduction band, where electrons are able to jump from one band to another.

The conductivity of intrinsic semiconductors is strongly dependent on the band gap. The only available carriers for conduction are the electrons which have enough thermal energy to be excited across the band gap.

Band gap engineering is the process of controlling or altering the band gap of a material by controlling the composition of certain semiconductor alloys, such as GaAlAs, InGaAs, and InAlAs. It is also possible to construct layered materials with alternating compositions by techniques like molecular beam epitaxy. These methods are exploited in the design of heterojunction bipolar transistors (HBTs), laser diodes and solar cells.

The distinction between semiconductors and insulators is a matter of convention. One approach is to consider semiconductors a type of insulator with a low band gap. Insulators with a higher band gap, usually greater than 3 eV, are not considered semiconductors and generally do not exhibit semiconductive behaviour under practical conditions. Electron mobility also plays a role in determining a material's informal classification.

Band gaps depend on temperature because of thermal expansion. Band gaps also depend on pressure. Band gaps can be either direct or indirect bandgaps, depending on the band structure.

http://www.wordwebonline.com/search.pl?w=band+gap

We've had light-emitting diodes and lasers for more than 40 years, transistors for almost 60. But until now, we haven't had a single device that can take an electrical input and simultaneously output both an electrical signal and an optical signal.

The emergence of the transistor laser has taken a long time, but it's not because of lack of interest?but because of a dearth of ideas. Researchers have measured light emission from transistors before. In the early 1980s, a research group at the California Institute of Technology, in Pasadena, led by graduate student Joseph Katz even fabricated a few experimental devices they called translasers. Using a wire, they integrated a transistor with a laser diode to fashion a device that could produce both electrical signals and laser beams, though not both simultaneously.

http://spectrum.ieee.org/feb06/2800

Roughly stated in these amazing articles is the very foundation for the future of optical computation. The hurdles and solutions found thus far in the industry bring us optical transceivers from both IBM and Intel as well as a host of companies involved in networking and server Opticals.

we at Intel's Photonics Technology Lab have been working on these building blocks for several years. One of our latest achievements, announced last February, is the world's first continuous all-silicon laser, which is based on the Raman scattering effect. Named for the Indian physicist Chandrasekhara Venkata Raman, who first described it in 1928, this effect causes light to scatter in certain materials to produce longer or shorter wavelengths.

Raman scattering is used today, for example, to boost signals traveling through long stretches of glass fiber. It allows light energy to be transferred from a strong pump beam into a weaker data beam. Most long-distance telephone calls today benefit from Raman amplification.


Typically, a Raman amplifier requires kilometers of fiber to produce a useful amount of amplification, because glass exhibits very weak scattering. Silicon, though, has a crystal structure, so its Raman scattering is more than 10 000 times as strong as that of ordinary glass fiber. In other words, you could achieve the same amplification in a centimeter-square chip that you'd get in kilometers of glass fiber.

In fact, so intense is the light amplification in silicon that it sets the stage for creating a laser. To build a Raman laser in silicon, you first need to create a conduit, also known as a waveguide, for the light beam. This can be done using standard CMOS techniques to etch a ridge or channel into a silicon wafer. In any waveguide, some light is lost through imperfections, surface roughness, and absorption by the material. The trick, of course, is to ensure that the amplification provided by the Raman effect exceeds the loss in the waveguide

http://spectrum.ieee.org/oct05/1915/3]IEEE Spectrum: The Silicon Solution[/url]

With silicon, some of these extra assembly steps can be greatly simplified by making them part of the wafer fabrication process. For example, the ends of the chips can be etched away to a mirror-smooth finish using a procedure known as deep silicon etching, first developed for making microelectromechanical systems. This smooth facet can then be coated with a dielectric layer to produce an antireflective coating.

Because a fiber and a waveguide are different sizes, a third device?typically a taper?is needed to connect the two. The taper acts like a funnel, taking light from a larger optical fiber or laser and feeding it into a smaller silicon waveguide; it works in the opposite direction as well. Obviously, you don't want to lose light in the process, which can be tricky when hooking up a waveguide 1 micrometer across to a 10-mm-diameter optical fiber.

Connecting optical fibers to optical devices on a chip requires attaching the fiber directly to the chip somehow. One approach we are pursuing is micromachining precise grooves in the chip that are lithographically aligned with the waveguide. Fibers placed in these grooves fall naturally into the proper position. Our research indicates that such passive alignments could lose less than 1 decibel of light as the beam passes from the fiber through the taper and into the waveguide.

http://spectrum.ieee.org/oct05/1915/soluf1

http://spectrum.ieee.org/oct05/1915/3]IEEE Spectrum: The Silicon Solution[/url]

Silicon is a lousy light emitter. To understand why, you need to know something about its electronic energy structure. In a typical semiconductor, the regular, repeating arrangement of atoms in its crystalline form results in distinct bands of closely spaced energy levels; these are the allowable energy states of the crystal's electrons. In between those bands are gaps where electrons cannot exist. For most practical purposes, only two bands really matter: the valence band, which contains the energy levels normally occupied by electrons, and the band immediately above it [see illustration, "Mind the Gap"]. The upper band is called the conduction band, because electrons energetic enough to reach it become mobile and free to accelerate under the influence of an electric field, thereby constituting an electric current. The difference in energy between the top of the valence band and the bottom of the conduction band is known as the band gap.


Normally, electrons occupy the valence band, but give them the right dose of heat, light, or voltage, and they will jump to the conduction band, leaving behind something called a hole, which is basically the absence of an electron in the crystal lattice. However, this electron/hole pair?an exciton?is a fleeting thing; sooner or later, the electron falls back to the valence band and recombines with a hole. Because energy is always conserved, this recombination of an electron and a hole is accompanied by the emission of a particle, preferably a photon, whose energy matches the difference between the conduction band and the valence band?the bandgap energy.

Energy, however, is not the whole story. Electrons also have momentum, and when an electron/hole pair is created?or destroyed by recombination?both energy and momentum are conserved. In direct-bandgap semiconductors, such as gallium arsenide, it happens that the maximum energy in the valence band and the minimum energy in the conduction band occur at the same value of electron momentum. With these direct-bandgap materials, an electron that has been excited into the conduction band can easily fall back to the valence band through the creation of a photon whose energy exactly matches the bandgap energy. Photons lack momentum, so it's a straight swap: all the energy of the bandgap jump goes into the photon.

http://spectrum.ieee.org/oct05/1886/2]IEEE Spectrum: Light From Silicon[/url]

http://spectrum.ieee.org/oct05/1886/3]IEEE Spectrum: Light From Silicon[/url]

SALVATORE COFFA is deputy director of STMicroelectronics NV's microcontroller, linear and discrete group, based in Catania, Italy.

Our approach combines both techniques. It has produced light emitters that operate at room temperature with a controllable tradeoff between high efficiency and long lifetime. The device structure looks very much like the metal-oxide-semiconductor transistors that make up the circuits in most microchips [see illustration, "The Silicon LED"]. Atop a region of p-type silicon we built a thin insulating layer of what's known as a silicon-rich oxide. That's simply silicon dioxide with a little extra silicon. Rare-earth ions are implanted in the middle of the oxide layer, and it is heated. The heat causes the silicon to clump spontaneously into crystals a few nanometers across. To finish the device, an n-type silicon layer is added with a metal electrode on top of it.

Applying a voltage to the electrode sets up an electric field that accelerates electrons across the silicon-rich oxide layer. These "hot" electrons collide with the rare-earth ions, kicking them into energy states that lead to light emission. The silicon nanocrystals have two roles. First, they greatly improve the conductivity of the silicon dioxide layer, and that boosts the device lifetime, though it reduces efficiency. Second, instead of emitting light themselves, the nanocrystals act like energy funnels leading to the ions. Hot electrons or emitted photons excite the nanocrystals, which then transfer their excitation to nearby ions, adding to the light emission.

The result is a device that glows brightly at room temperature, with a quantum efficiency of up to 10 percent, comparable to that of state-of-the-art III-V devices.

A great advantage of the technology is that the color of the light emitted depends only on the rare-earth ions used. Samarium glows red; terbium, green; cerium, blue; and erbium, conveniently, yields the infrared used in many telecommunications devices [see photo, "Color Codes"].

The problem with the approach for some uses, at present, is low light output. Although the silicon LED can be as efficient as its III-V competitors, it produces only a fraction of the light you'd get from a commercially available LED. That's because the maximum output power is limited by how densely we can pack the device with rare-earth ions, and that limit at the moment is about 1 quadrillion ions per square centimeter.

To get from LED to VCSEL in a resonant cavity, the light bouncing between the mirrors must manage to excite the majority of the rare-earth ions, creating a state called population inversion. The performance characteristics of the erbium-based, silicon-rich-oxide LEDs we've already built indicate that achieving population inversion is within reach. All we need to do is further reduce the amount of light lost within the structure to imperfections in the mirrors and elsewhere. We're confident we can accomplish that within a couple of years.

So get ready for light-speed links to your PC and even inside it. An electrically driven silicon laser is just around the corner.

http://spectrum.ieee.org/oct05/1886/3]IEEE Spectrum: Light From Silicon[/url]

IBM optical roadmap

IBM chip

Intel Terachip.

i certainly hope that I have illuminated alot of the hurdles and solutions , the brilliant minds at STM and Intel have been working on for over a decade and to everyones surprise we see a tech from 1928 bieng used in some forms of Intels technologies; That is simply amazing. By the IBM roadmap there isnt much to carryover for a typical user on this optical monolith; There are graphene and carbon technologies to assimilate as well which may come into play in future on chip optics.
Interestingly we see a flood of recent acceleration technologies even on silicon based chips, but still I think the future will bring light and sooner than later.

i would like to see AMD make an optical northbridge/imc/southbridge adaptation, and intel utilize the silicon laser in a nb/sb configuration. And ultimately I would like to see pci and pcie replaced by optical connections as well as all of the I/O in time sound bieng the first to switch interfaces.

While on die solutions are well ahead of any current or immediate enthusiast fixations, I feel they are within the planning range; If you hadnt noticed there is alot of tech at a stage nearing practical use, we even have the flexible screens i mused about in my short RFID story back in february. Back then we had talked about flexible circuitry and I had said to others disbelief that rollup laptops would arrive some day.

The future of tech will always be fast flexible and multi tasked, each gen will arrive at better solutions until a watch can contain todays power and the wifi that links it is faster than todays dsl. Ultimately the power needs to be small discrete ,even fashionable and mega powerful, I could easilly see cellphone form factors replacing laptops some day.

For now this should conclude part one. cheers dewd.

http://www.eetimes.eu/20000064...U52QSNDLRCKHSCJUNN2JVN]EU researchers demo Si-based optical interconnects[/url]



John Walko
EE Times Europe
06/26/2007 11:33 AM

LONDON ? European researchers have demonstrated an integrated device that could form the basis for on-chip optical interconnects. The group integrated electrically-pumped microdisk lasers with a nanophotonic silicon wire waveguide.

The demonstration is the first outcome of a three year collaborative research project that started in 2004 with a budget of 4.2 million euros (about $5.4 million) and the goal of devising methods to integrate optical interconnect on top of silicon ICs.

Dubbed PICMOS (Photonic Interconnect layer on CMOS by wafer-scale integration), the project linked ST Microelectronics and Interuniversity Microelectronics Center (IMEC) among its partners. Others include the University of Ghent in Belgium, the Grenoble-based CEA/LETI laboratory and Tracit Technologies, the University of Lyon and the Technical University of Eindhoven.

Markus Riester

A paradigm shift in production methods for optical waveguides could make the technology more cost-competitive.The problem of how to embed optical connections in printed circuit boards (PCBs) has been an area of intense investigation since the 1980s.1 Approaches to solutions include materials, process technology, integration concepts, and connecting active components to waveguides. These methods generally involve cladding/core/cladding structures and a series of processing steps, independent of the details of the waveguide configuration. Although the resulting optical performance is generally adequate, all the solutions suffer from the strong competition of electrical waveguides, which are commonplace wherever wiring is used to connect electrical components.
In view of the anticipated need for optical waveguide technology, its commercial success has so far been limited. Although the technical feasibility is undisputed, customer applications determine the critical parameters for making the right choices for one or the other solution. Aside from performance standards, one particularly critical requirement is straightforward integration of old and new technologies without the need for major or even minor changes to the existing electrical design. While this is easier said than done, it is likely to be one of the success factors for the implementation of optical PCB technology.

http://spie.org/x14992.xml?highlight=x2402]Integrated optical interconnections on printed circuit boards: Newsroom: SPIE.org[/url]

Broadband optical amplification on a silicon chip

Alexander Gaeta and Michal Lipson

Four-wave mixing can be used to design on-chip amplifiers with unprecedented bandwidth and low power.
Chip-scale photonics are becoming increasingly promising for replacing some of the copper interconnects in con
ventional microelectronic chips for applications that require low power and high bandwidth. Several components compatible with current silicon microelectronics have already been demonstrated, such as highly compact electro-optic modulators,1 all-optical switches,2 low-loss
waveguides, and filters.

A critical component that remains to be developed is an on-chip amplifier. Although the indirect bandgap of silicon has made it difficult to create such an amplifier based on stimulated emission, researchers have recently exploited the large effective optical nonlinearities of silicon to produce an alternative amplification mechanism. The first demonstration of the use of a nonlinear process to produce amplification was based on the Raman effect.3 However, the Raman bandwidth over which amplification occurs is relatively narrow, which prevents its use in wavelength-division-multiplexing systems that require gain at least over tens of nanometers.

http://spie.org/x14751.xml?highlight=x2402]Broadband optical amplification on a silicon chip: Newsroom: SPIE.org[/url]


Optical logic gates based on micro-electro-mechanical systems

Gustavo Pamplona Rehder, Marco Isaías Alayo Chávez, Hector Baez Medina, and Marcelo Nelson Paez Carreño

Optical information conducted through integrated waveguides fabricated on silicon substrates can be processed using AND and OR logic gates by combining micro-switches.
The continuous push to increase processing performance of electronic components has, until now, been achieved by a constant reduction of the transistor geometry in metal-oxide-semiconductor (MOS) technology. The increased performance of integrated circuits (ICs) is also associated with the miniaturization of and increase in packing density (i.e., the number of transistors in a single chip), which reduces cost. This, however, also increases the power dissipated as heat, which can be detrimental for correct operation. These physical limitations of the current technology have lead us to search for alternative ways to process information.
The integration of optics with electronic devices using silicon technology is an alternative that offers faster processing speeds due to the wider bandwidth of optical signals. Further, devices based on optics have lower electromagnetic sensitivity and a smaller thermal budget.1,2 Moreover, the development of these electro-optical devices follows a natural path, since optical fibers are

http://spie.org/x14986.xml?highlight=x2402]Optical logic gates based on micro-electro-mechanical systems: Newsroom: SPIE.org[/url]

ALMOST A LASER: A resonant-cavity light-emitting diode is similar in structure to a laser and comes close to producing a laser's monochromatic light [inset]. It is made by embedding the layer of silicon nanocrystals and rare-earth ions from a silicon LED between mirrors consisting of alternating layers of silicon and silicon dioxide. Light from the nanocrystal layer bounces back and forth between the mirrors, stimulating the emission of even more light with each pass, until it finally exits the device

While not everyone may not see eye to eye on this issue i think the sources have done a good job of laying out the facts regarding this development. And it is always a benefit to have great discussion on such topics.