Intel Corporation 1968 - 2003

Below is a summary of the HBS case called Intel Corporation: 1968 - 2003

The first section talks about how intel invented the microprocessor in 1971. I found a link on about.com that has more info if you're interested.

http://inventors.about.com/od/mstartinventions/a/microprocessor.htm

The section talks about how intel was at its peak in the summer of 2000, with its stock valuation at $510 billion. It goes on to talk about a down turn in the semiconductor industry in December of 2002. Basically all Intel's competitors scaled back R&D and manufacturing capacity. Intel's CEO, Craig Barrett decides that innovation is going to be the key to getting out of the downturn. So he invests more money into R&D and manufacturing. Intel also invests heavily into silicon for networking and communications (think Internet). The author basically questions if this bet is going to pay off.

Intel basically starts with two engineers that leave a company called Fairchild Semiconductors. One of the engineers is Gordon Moore (think Moore's law). They also get 2 million from a venture capitalist. They begin the process of gathering up all the smart people from Fairchild.

The author then talks about Intel as a memory chip company. Basically intel first starts out by creating the worlds first semiconductor memory device. This replaces magnetic core memory. Intel tried 3 designs, but eventually went with MOS, metal oxide semiconductor transitors. Intel created the worlds first 1kb DRAM.

The author then goes into how there was a network of patent sharing, since one design might need 100 patents, companies agreed to share the patents. Since Noyce (one of the founders) knew a lot of people, Intel was able to get in on the patent sharing even though it was a new company. The author also talks about how expensive it is to start up a chip company. These two factors limit potential entrants in the chip business.

There's a section about competition from Japan. Basically Japan had easier access to capital through bank loans and had a more efficient manufacturing process. They turned out less defects, and worked with suppliers of equipment to make the manufacturing process better.

Every time a new DRAM chip came out, Intel would lose more and more to their competitors. There was the 16k and 64k dram chips. Intel finally changed the DRAM business again by introducing the first CMOS DRAM chip at 256k, but by the time they came out with the 1 meg chip, the intel lost the business relevance of dram.

EPROM - Intel talks about how they really made their fame off of DRAM, but their money came off of EPROM. Eprom, was easily programable. Basically if you made a mistake with regular ROM you'd have to throw it out or start from the beginning and burn/etch everything again. With eprom you just flashed UV light on the circuits and that wiped them out. Intel made a bold move of increasing the price of eprom 100 fold. They made a killing. For a long time intel had no competitors in the eprom game as they flew under the radar. When the Japanese finally came into the game, prices dropped 90%. Intel also invented FLASH memory.

Intel invents the microprocessor in 1971, but they just saw it as putting some chips together. A couple of Intel scientists recommend it for personal computers, but intel shoots it down. Eventually in 1977, radio shack and commodore enter the pc market. Intel's chip goes into the Altair computer kit. Apple ends up using Motorola's chip in the apple II pc.

Intel's DRAM business is hemorrhaging. There are way to many competitors, and intel has only a 1% market share. Memory is now losing money for the company. They enventually decided to get out of the memory business, but stay in the flash memory game.

In 1981 IBM enters the PC market and basically uses an open standard, this is the big push in the microprocessor industry. Intel and motorolla will now battle it out. Intel wins the IBM contract. But, IBM forces intel to license it's microprocessor to other companies, so the IBM can ensure having supplies from more than one company. One of the companies is AMD.

Intel invents the 80386 chip, aka the 386 chip. It tells IBM that it will not let other companies manufacture it. It will be the sole provider. IBM gets scared since that the chip will actually take money away from its midrange computers (AS400 type). Compaq comes out of nowhere and starts making the first 386 computers. 386 chips sell for much more money. It's the first time and IBM clone held the lead in performance.

AMD ends up suing intel for rights to manufacture the 386, and then even the 486. Intel makes an agreement that AMD can sell the 486, but that's it. They have no rights to anything in the future. Including the famous pentium. For those who don't know, the pentium chip is basially what put intel into the stratosphere. That was the chip that everyone had to have in their computers.

RISC vs CISC, in microprocessor land, this was VHS vs BETA all over again. Intel based their architecture on CISC, complex instruction set computing. RISC was based on reduced instruction set computing. Back then programs where written at the chip level, (assembler language). RISC had it's own language, and so did CISC. Intel eventually copied some ideas form RISC into it's chip and customers didn't want to rewrite their x86 programs for just a little bit of performance gains.

Intel inside. A good analogy to the intel inside sticker is like an M3 emblem on a BMW, but imagine if that M3 emblem could be on other cars. Intel basically gave money back to the computer manufacturers so that they would advertise intel inside, and also put the now famous sticker on their pc's. 300 manufacturers jumped at the chance. IBM's advertising agency didn't like the idea.

In 2000 there was less willingness by consumers to pay for the "brand".

At one time intel even sold pre built pc's to Prime, AT&T, and Unisys. They eventually exited the low margin business mainly from pressure from OEM's.

Intel priced new chips high on purpose so that they could limit demand. Once they lowered the cost of the chip they then lowered the price up to 50% withing a year.

The pentium problem which would only affect those who did heavy duty mathematical computations cost intel 475 million dollars. Intel eventually succumb to pressure and decided to honor all chips that were sent back to them.

Suppliers, intel pushed their suppliers to give them the most cutting edge products first.

Complimentors- Microsoft was one of the complimentors. Historically hardware was always ahead of software. Intel even paid game designers money to use the MMX instructions int their pentium chip.

Intel capital -- intel formed intel capital to specifically invest in businesses that could help intel. They made billions of dollars on it.

Intel was sued by the FTC over anti-trust, but nothing real came out of it.

Barrett their new CEO see the writing on the wall that the microprocessor will no longer be the big money maker. He sees the pc being overtaken by the cell phone, laptop.... He changes intel's mission statement to talk about networks, wireless, communication, and the Internet.

He reorganizes the company into 4 different groups.

Client Platforms -- include desktop and mobile processors.

Intel takes heat from AMD. AMD buys the atahlon chip and introduces the 1 gig chip. Intel fights back by putting a chip in every price category. Celeron is their cheap chip.
In the wireless game intel focuses on improved battery life, and reduced heat. They redesign their chips. They can do this because corporate customers don't care about price so much. Intel is finally changing their strategy of competing on performance.

Server Platforms -- Xeon is their 32 bit processor. Itanium was their 64 bit processor. Itanium ran existing software slower, because of the architecture change. Also people have to rewrite their code to utilize the 64 bit platform. It gained small market share. Out of left field AMD came to the game with their own 64 bit chip for servers. This chip ran existing 32 bit software better.

Cellular and Wireless -- Intel develops XScale. Their main competitor is Texas instruments. Intel only started with 8% of the market share and TI had 14%. Intel did well with XScale. Intel pushed WiFi and is big in that. While Nortel, Lucent, and Alcatel fired people, Intel continued investing in communications.

Intel ran into some technological problems. They invested heavily in RDRAM and required it for the chips. That blew up in their faces. They also missed some deadlines and were short on supplies. Barrett initiated a back to basics campaign. It worked. Intel wants to be the building blocks of all Internet communication. At the end of the paper Intel selects a new COO so that Barrett can focus on strategy.