Replies to post #15870 on Intel Corp (INTC)

[Saturn V]

You are missapplying the term "all optical chip". The Intel invention is still about electrical computing with an optical interface,instead of the traditional electrical interface.

Light is capable of carring millions of gigabits of information per second. However todays electrical modulators of a light source can only switch at several Giga Hz per second, and these electrical control signals are the bottleneck for transmission. (Modern optical fibers carry more data because they transmit several light beams of different wavelengths.) Researchers have dreamt of eliminating the electrical bottleneck by all optical switching, ie by using one light beam to modulate another. In principle such an all optical system could carry millions of GB of information per second on a light beam. Unfortunately no decent optical switching element has been developed.

The Intel development is not about all optical computing. Today the bottleneck in computing is memory access. Acess to DRAM memory takes several hundred clock cycles, because the electrical interchip connection is very slow. On chip speeds are very fast since the capacitive loads are tiny. So a CPU interfacing to memory or other components via an optical fibre would alleviate the memory bottleneck. However this is not an all optical system. The system is still an electrical switching system, but in which certain high speed chip interconnection is accomplished via optical means.

Yes if optical interchip connection becomes viable, computer architecture will change. Today magic mantra of maximizing performance via ever larger on chip cache memory may have to be rethought. Todays slow electrical interface has caused everyone to pack more components on one chip to bypass the slow I/O interface. Optical interface might change that and might bring back the world of large computing mainframes.

[Saturn V]

Hi Mike,
Here is link on new technique of Optical modulation of Lasers by Si-Ge chips. The work was done at Stanford and funded by Intel.
http://www.nytimes.com/2005/10/26/technology/26cnd-chip.html
Engineers Report Breakthrough in Laser Beam Technology


SAN FRANCISCO, Oct. 26 - A team of Stanford electrical engineers has discovered how to modulate, or switch on and off, a beam of laser light up to a 100 billion times a second with materials that are widely used in the semiconductor industry.

The group used a standard chip-making process to design a key component of optical networking gear potentially more than 10 times faster than the highest-performance commercial products available today.

The team reported its discovery in the current issue of Nature, which was published on Wednesday. Such an advance could have broad applications both in accelerating the already declining cost of optical networking and in potentially transforming computers in the future by making it possible to interconnect computer chips at extremely high data rates.

Currently, the communications industry uses costly equipment to transmit data over optical fibers at up to 10 billion bits per second. However, researchers are already experimenting with optically linked computers in which components may be located on different sides of the globe. Cheap optical switches will also make it possible to create data superhighways inside computers, making it possible to reorganize them for better performance.

"The vision here is that, with the much stronger physics, we can imagine large numbers - hundreds or even thousands - of optical connections off of chips," said David A.B. Miller, director of the Solid State and Photonics Laboratory at Stanford University. "Those large numbers could get rid of the bottlenecks of wiring, bottlenecks that are quite evident today and are one of the reasons the clock speeds on your desktop computer have not really been going up much in recent years."

The modulator, or solid-state shutter, reported by the team, could also have a dramatic effect on the telecommunications industry, which is already being transformed by the falling cost of optical fiber networks.

The device, which is constructed from silicon and germanium, would alternately block and transmit light from a separate continuous wave laser beam, making it possible to split the beam into a stream of ones and zeros.

The effect, known as a Quantum-Confined Stark Effect, or QCSE, has been previously demonstrated, but was not expected in the germanium, a material that is compatible with the industry's silicon-based manufacturing technologies.

The Stark Effect allows materials to act as shutters for particular wavelengths of light as an electrical field is switched on and off. In the past, however, the effect has been achieved in optoelectronic applications by using exotic materials like gallium Aarsenide, which are not easily compatible with standard chip-making techniques.

"What we achieved is somewhat surprising," said James S. Harris, a Stanford University electrical engineering professor, who is a member of the research group. "No one thought it would work."

The research project was supported both by the Intel Corporation and by the Pentagon's Defense Advanced Research Projects Agency. Intel has been intensely interested in the possibility of designing optical communications components with standard chip-making tools, both for networking and computer communications applications. Theodore I. Kamins, a quantum materials specialist at Hewlett-Packard Laboratories, also contributed to the research effort.

"They've made a big leap," said Mario Paniccia, director of Intel's Photonics Technology Laboratory. That research group has made a number of announcements about progress in development of similar components that could lead to low-cost optical network systems in the future.

He acknowledged, however, that there is a significant gap between research results and commercial availability of devices based on scientific breakthroughs.

Other designers working in the field were also cautious about direct applications of the technology. Alex Dickenson, chief executive of Luxtera, a Carlsbad, Calif. start-up firm that announced a 10-billion bit per second optical modulator using a different silicon-based approach earlier this year, said that he believed there would significant hurdles to the commercialization of the Stanford discovery.

He cautioned that while the display was interesting from an academic perspective, the researchers had yet to prove that the effect works at the standard frequencies of light used by the telecommunications industry.

Several industry executives said the advance was significant because it meant that optical data networks were now on the same Moore's Law curve of increasing performance and falling cost that has driven the computer industry for the past four decades. In 1965, the Intel co-founder Gordon Moore noted that the number of transistors that could be placed on a silicon chip was doubling at regular intervals. The semiconductor industry has held to that rate of change since then, giving rise to the modern era of microelectronics that has transformed the global economy.

Now that rate of change could be directing the future of the telecommunications industry. Computer and communications industry executives believe that advancements in inexpensive optical networks will transform the computer industry and other major industries ranging from the financial marketplace to Hollywood.