Aiming to become the global leader in chip-scale photonic solutions by deploying Optical Interposer technology to enable the seamless integration of electronics and photonics for a broad range of vertical market applications

Reply New Message

Prev Message Back To Forum Threaded View Next Message

maing

Rank: Treasurer [?]

Points: 3480 [?]

Votes: 136 Score: 2.4 [?]

Your Vote:

Please Log In to Vote

Did you know? You can earn activity points by filling your profile with information about yourself (what city you live in, your favorite team, blogs etc.)

German Article in English... Excellent

posted on Feb 02, 2014 09:34AM

Rainers Blog

Recent Posts

>
Enter your email address to subscribe to this blog! For new posts, you will receive an e-mail.

Team up to 43 followers

POET Technologies: High-speed digital technology of tomorrow

Written by Rainer Klute - 2014-02-01

POET Technologies is a Canadian company that has developed a manufacturing process for fast, low-power, low-cost optoelectronic and semiconductor chips based on gallium arsenide. POET wants to go beyond the limits, before the silicon semiconductor industry today.

Figure 1: transistors and clock speeds of Intel processors (Source: Herb Sutter )

If you ever noticed that computer for ten years no longer faster?

Moore's Law

1985 When I bought my first own computer - an Atari ST - had a CPU clock speed of 8 megahertz (MHz). In the following years the speed of new processors increased rapidly and now stands at 3-4 gigahertz (GHz), which is the 500-fold of my old Atari.

The Moore's law describes this rapid growth. Gordon Moore postulated in 1965, the number of transistors per unit area in integrated circuits will double every eighteen months to two years. Nearly fifty years, agreed that too, like the green line in Figure 1. Today, however, the Moore's Law has reached its limits, because you can the structures of the chips of course not be arbitrarily reduced.

Less is not more expensive

The smallest physically possible structure size is not yet reached, but there is also an economic limit. Smaller structures requires significant research and development effort, more complex manufacturing process and a total of significantly higher costs. Modern chip factory can cost several billion dollars. Eventually, a reduction is still possible, but they no longer pays. This point is reached now, as Figure 2 shows.

Figure 2: Number and size of transistors (source: The Economist )

While you got in 2002 for a dollar 2.6 million transistors having a size of 180 Namometern (nm), there were 2012 20 million transistors with 28 nm 2014 shrink the transistors on 20nm. However, they are no longer cheaper. 2015, the trend is reversed then completely around: Due to the high development and production costs further scaled structures are transistors for Use same money is then given less.

Meltdown

While the structure size as in smartphones plays a role because of the space required, it has long been a matter for computer processors. Because since 2003, the CPU clock speed already stagnating at about 3 to 4 GHz (Fig. 1, dark blue line). Higher clock rates have a higher power consumption and higher heat generation result. This heat can no longer dissipate quickly enough though. The structures of the processors would melt, and the finer the structures, the more likely they melt. So you raise the clock speeds not share, but leaves them where they are. For performance improvements can be used instead or many processors with multiple CPU cores.

The heat generation is of course also in smartphones is a problem. Unpleasant is not only when the device is warm after some operations. Even more uncomfortable it that even the battery is then quickly empty.

Rotten moor esches law

The physical and economic end of Moore's Law is a huge issue for the semiconductor industry. How to sell chips that are neither faster nor cheaper, more fuel efficient? Chips, which are not even less in the near future? Trying all manner, as chip designs with three-dimensional structures, or a cooling liquid metal circulating in special ducts through the chip. However, a major breakthrough is in sight.

High speed, low power consumption

This problem now promises to solve a small company in Canada. Instead of the conventional silicon (Si) as the base material of the chip manufacture, is POET Technologies on gallium arsenide (GaAs). Dr. Geoffrey W. Taylor, chief scientist of POET Technologies has made it in about 20 years of research to solve the problems of this material, where in the 1990s a number of companies have failed. POET is Taylor's life's work.

Using the POET platform (POET stands for P Lanar O pto E lectronic T echnology) can be chip with the following properties can be produced:

Significantly higher speed
Significantly lower power consumption
Integration of electronics and optics on a chip (Optoelectronics)
Integration with conventional silicon technology possible
Manufacture of the POET chips on existing installations possible

At the moment, POET is to reduce the feature size to 100 nm. Not so easy on your own old system! Up to 200 nm one has succeeded now. How POET announced , is expected at 100 nm with at least 50-fold speed compared to today's silicon chips. Depending on the application, the power consumption should be between 10 to 25 percent. And that's just the beginning, because on modern systems can be still much smaller structures finished.

No new production

This brings us to another not to be underestimated advantage of the POET process: Chip manufacturers do not need new equipment, they do not need billions of dollars to invest. POET chips can be fabricated on the conventional systems. Low investment in retooling rich.

Combines optics and electronics

POET enables the integration of electronic and optical functions on the same chip. Therefore in order to that which today's devices require several interconnected chips in a single implement. This reduces the manufacturing costs. Devices "powered by POET" are not only faster and more economical, but also cheaper. Future applications of optoelectronics give an idea at best.

Exciting is the high radiation resistance of GaAs. This finds the NASA wants to use the POET technology in a space probe. In space radiation resistance is so important.

More than smoke and mirrors

Too good to be true? Concrete projects POET Technologies has performed for and with BAE Systems, show: this technology exists and works. BAE currently validated the manufacturing process for some optical components. Or better: Probably the validation has already been completed, with the announcement of the achievement of "milestone 7" is expected daily. These POET but relies on BAE.

Marketing

Currently POET Technologies is going to make the technique to money. For this purpose we speak since 2013 behind very firmly closed doors with potential licensees or potential partners for other types of partnerships. What will be discussed with whom, is unknown and runs under strict secrecy. But apparently is from the structural reduction to 100 nm, the requirement of at least one potential partner. For this aim originally was not on the roadmap, but came in August 2013 added surprising.

Possibly also at the suggestion of a potential partner to get for now distinguishes between POET and PET chips. The latter would be without the optics, rather with conventional Si chips comparable and therefore probably faster to market.

While 2012 was a sale of the company mentioned, this variant now appears to be off the table. I like that, because the POET process is too revolutionary to leave it to a single manufacturer, and thereby give him a monopolistic market advantage. It is better to license the technology to manufacturers would be. And POET is interesting not only for computer and smartphone processors, but also for numerous other vertical markets. Therefore, I hope to licensing agreements with as many manufacturers as possible from many areas. This results in a steadily flowing, ever-increasing stream of royalty payments would generate.

The POET-share

Whoever finds it all interesting, can buy POET Technologies shares and so share in success or failure.

Personally, I see a huge potential in this technique and have therefore invested some money in POET shares. That may tarnish my view. Nevertheless, I wish to describe my thoughts short.

Currently, the market capitalization of POET Technologies at a share price of 0.75 CAD is (2014-01-31 closing price on the TSX Venture Exchange ) and nearly 135 million shares its just over $ 100 million (67 million euros).

What is POET Technologies really worth it? If it goes to the usual standards, nothing at all, because:

The financial structure is structurally very bad. POET Technologies has a pure R & D firms virtually no revenue, but burns month the existing chapter.
The POET-stock is a penny stock. That is, they listed under a dollar and with that comes for virtually all institutional investors in principle not considered.
Professional analysts have POET Technologies not yet been looked at. So there are no estimates on which investors could be emulated.
The course is very volatile, and subject to substantial fluctuations and not for the faint of heart.

This one is POET Technologies course not fair. The true value of POET's not plugged into the current, but in the expected future financial data. And I think - especially in comparison to what is otherwise paid for companies in the IT sector - that for POET Technologies roughly a rule of thumb at least $ 10 billion likely to be inside. Too high, this value is probably not, but will rather be at the lower end of the flagpole. But I reckon rather conservative!

Which share value is derived from it? Unfortunately, you can not divide by today's 135 million shares, the $ 10 billion. Because this number is expected to increase by outstanding options and warrants on some 212 million shares. It will therefore result in a corresponding dilution. We therefore share 10 billion dollars by 212 million shares and arrive at around 47 dollars a piece - a Versechzigfachung compared to the present value.

POET shares are listed on the TSX Venture Exchange listed in Toronto (Canada). At the German stock exchanges they are traded under ISIN CA73044W1041. On most liquid are Tradegate and Frankfurt, you can also via Tradegate 8-22 clock act. This fits well with the trade in Toronto, as this will take place from 15:30 bis 22:00 clock clock our time.

Disclaimer

Not that anyone misunderstands: I advise anyone here to buy POET shares. The thought of me $ 10 billion are my own, highly personal conjecture. Also for the run from the POET-management negotiations, there is no guarantee of success. So if anyone of you is shipwrecked with POET stocks, you please do not complain to me! Informs you rather own, thoroughly reviewed, and then decide.