as the company which did the original design puts a chip on the market, the chip pirate purchases it, removes any impeding coatings on the chip surface, and sends it to a photomicrographics specialist to make blow-up photos of the layout design. Typically, a blow-up 800-1000 times the original chip size is used. The chip pirate (or a commercial business which offers chip copying as a service) electronically traces the photographic blow-up and feeds the design information into a computer in exactly the same way as the original digitizer did from the original layout drawings. The techniques and the equipment are exactly the same.

To demonstrate that the chip pirates do exist, I have a slide of one of my company's products-the 16K Ramdom Access Memory.

Now if direct copying of an integrated circuit is not an available alternative, the chip pirates cannot exist. These circuit designs are so intricate and complicated that any attempt to make geometrical changes in the layout just for the sake of change will surely meet with disaster. The only way this can be done is for the copier to gain a thorough technical understanding of the circuit so that an alternative layout can be made to work. This requires a lot of patient study by sophisticated engineers and is known as "reverse engineering". Oftentimes reverse engineering can consume nearly as much time and effort as the original development did, and oftentimes reverse engineering produces an improved product over the original. We have no quarrel with that. It's fair game.

Line-by-line copying, however, is quick, cheap, and allows a competitor unfair advantage by drastic cutting short development time and expenses. For example, our company spent over two years and over $3 million developing the 16K RAM to the producible product shown in an earlier slide. There is a company in Japan that can be hired to copy it in less than three months for less than $50 thousand. The proposed amendment to Section 101 of the Copyright Act of 1976 will provide our industry protection against this sort of thing.

Thank you for this opportunity.

TESTIMONY OF ANDREW S. GROVE, PRESIDENT, INTEL CORP., ACCOMPANIED BY ROGER BOROVOY OF INTEL

Mr. GROVE. Mr. Chairman, I would like to submit through evidence the written version of this testimony which you, I believe, have in your hand.

Mr. KASTENMEIER. Without objection that testimony you refer to will be received and made a part of the record.

[The written text of Mr. Grove's statement follows:]

STATEMENT OF ANDREW S. GROVE ON BEHALF OF THE AMERICAN ELECTRONICS

ASSOCIATION

Andrew S. Grove received his Ph.D. in Chemical Engineering from the University of California at Berkeley. Dr. Grove is the author of a leading text on semiconductor physics. This week, he will be installed as President of Intel Corporation.

AEA (formerly WEMA) is a trade association representing more than 1,000 hightechnology electronics companies in 39 states. While some of our member companies are among the largest firms in the United States, the majority are smaller businesses employing fewer than 200 employees. Most of our member firms design and manufacture sophisticated components and equipment for a number of end markets.

HIGH TECHNOLOGY INDUSTRIES DEPEND ON AGGRESSIVE R. & D. SPENDING

The high-technology electronics industries, of which the semiconductor industry is an integral part, play an increasingly important role in the U.S. economy. Our products are being used in a wide and growing variety of business, military, scientific and consumer applications that, by extending the powers of the human body and intellect can improve productivity and the quality of life. Our industries provide an ever-growing number of jobs and exports. A 1977 AEA survey of 325 of its member companies showed employment of 750,000 people in the U.S. In 1977 high-technology consumer electronics-oriented exports were $8.92 billion (source: U.S. Department of Commerce).

This industry was created, and its growth powered, by aggressive investment in research and development. Its future growth is dependent, if anything, even more strongly on continued heavy research and development spending.

It is in this context that I address H.R. 1007, which amends the Copyright Act of 1976. This bill will provide copyright protection to the semiconductor chip designs

and thereby maintain the incentive for firms to devote large resources to design and development.

SEMICONDUCTOR R. & D. COSTS SKYROCKET

One of the trends of the most advanced segment of the semiconductor industry is the exceedingly rapid growth of the cost of designing new products. This trend is shown in the first attachment, which is taken from the Keynote Speech at the 1979 International Solid State Circuits Conference, by Gordon E. Moore. It can be seen that while ten years ago a typical semiconductor integrated circuit took about ten person-months to design, the typical product today requires the expenditure of about 200 person-months-a 20-fold increase! Since the expense required for a person-month of design effort has also been steadily increasing, it is clear that the cost of semiconductor integrated circuit design has been-and will continue-growing at a dizzying rate. As a rough estimate, a 200 person-month effort costs in the neighborhood of $1,000,000.

No prudent management can authorize the expenditure of this kind of development sums unless the resulting product is protected from pirating by competitors desirous of taking a free ride. I shall provide some recent examples of this practice.

CHIP PIRATING ON THE RISE

The next attachment shows one of Intel's most important, advanced technology products-the 2147 4K static Random Access Memory (RAM). This product, the result of very extensive R&D work, enabled Intel's innovative Metal-Oxide-Silicon (MOS) technology to invade the marketplace formerly held by older bipolar products. IBM made their first purchase of memory systems from an outside supplier ever because of the unique characteristics of this memory component.

Intel introduced the 2147 in mid-1977 and has not, until recently, seen any competition. The first competition which recently appeared is a photographic duplicate manufactured in Japan by Toshiba. The attachment shows the Intel and Toshiba chips next to each other; clearly, the Toshiba device is a straight copy of the Intel device.

The next attachment shows a Russian copy of an Intel 4K dynamic RAM integrated circuit. To be sure, passage of the Edwards, Mineta and McCloskey bill into law in the United States will not prevent the Russians from copying in Russia. At least, however, it will prevent the Russians from exporting their illicit copies to the United States, and as Russian technology enters the 20th century it will become necessary for them, more and more, to conform their laws to ours to obtain exports. As I illustrated earlier, due to the extremely rapid increase in the cost of development of advanced semiconductor integrated circuits, the temptation for chip pirating, and the damage to the developer coming from such chip pirating, will inevitably increase unless protection is provided by the Congress. So far, companies such as Intel, Mostek, and other American semiconductor companies who spend huge sums on otherwise unsupported research and development have been able to reap the benefits of their ingenuity, and their shareholders, as well as our entire society, have been well served. But the inevitable rise of "chip pirating" does not augur well for the future. Chip pirates curtail the innovators' product leadtime (during which development costs can and must be recovered) by quickly reaping where the innovators have sown. If we lose the early profits from our designs to the chip pirates, funds available for development will be curtailed, and our industry will lose its technology lead. The semiconductor industry is a substantial exporter. We would suffer in balance of payments, stability of the dollar, and even in superiority of our military equipment as a result of loss of our semiconductor technology leadership position.

JAPANESE CHIP PIRATING PARTICULARLY DANGEROUS

If the pirating is done by the Japanese, the effect is doubly serious. The Japanese already have major competitive advantages: cheap money and a home market protected by tariff and non-tariff barriers to American exports. If we also allow them to help themselves to American technology by copying rather than having to do their own topographic designs, they will be handed the opportunity to take America's most successful high technology business away.

Patents are not enough to protect us. The Japanese make their share of patented "inventions." Since the American companies need licenses under the Japanese inventions in the same manner as the Japanese need licenses under the American inventions, patent exchanges have been the norm. On the other hand, to my knowledge, no American semiconductor company has ever made an unlicensed copy of a Japanese chip. If copyright protection is available, companies can exchange designs and receive a benefit back from a licensed copier.

In summary, if the semiconductor industry is to continue on its enormously successful path of providing jobs and exports, helping us conserve energy, and improving the quality of our lives in many other ways, it must have the assurance that its gigantic investment in research and development will have a chance to pay off: it needs protection from chip pirating!

300

200