How do you explain Intel’s initial dominance and subsequent decline in DRAMs?

Intel was founded in 1968 by Robert Noyce, co-inventor of the integrated circuits and Gordon
Moore, both having expertise in semiconductors after having worked at Fairchild. They recruited
Grove and many other people from Fairchild to set a team. Hence, Intel’s core competence was
building integrating circuits/ semiconductor devices.

There were following factors which drove Intel into producing DRAMs:

1. With the discovery of magnetic core memory in 1940 and growing demand of mainframe
computers, there was a huge growth potential in memory market.
2. Few semiconductor companies experimented with DRAM (Dynamic Random Access
Memory), a new type of high density memory that used integrated circuits but were unable
to mass produce them at lower cost.
3. Intel had a team with core expertise in Integrated circuits under the leadership of Robert
Noyce and this situation seemed to be a great opportunity for Intel to establish itself into the
market as a memory chip company.

The problem at hand was to mass produce DRAM at a lower cost and to solve this, Intel came up
with Metal-Oxide-Semiconductor (MOS) transistors, which was neither too easy for competitors to
copy nor too hard to build. Hence, Intel was first to produce 1-Kilobit DRAM using MOS process
technology which was not only cost effective but also boasted high performance and small build.
This gave Intel first mover advantage into the DRAM market which enabled them two years of
market dominance and generated huge cash flow.

Following were other factors that helped Intel in enjoying initial dominance in DRAM:

1. Complex Manufacturing process for DRAMs

2. Huge Fixed cost
3. One clean room (fab) required to extract maximum impurities from the environment
4. One expensive equipment to be replaced every 3 years
5. Increase in yield happened with troubleshooting
6. Chip price would reduce with high yield and capacity expansion
7. Access to sharing network for cross licensing of patented circuits to build a specific DRAM
deterred potential entrants
8. Cross licensing forced competitors to compete on Price/Performance

As per VRIO framework, Intel’s 1 Kilobit MOS DRAM was:

1. Valuable – Yes
2. Rare – No
3. Inimitable – No
4. Organized – Yes

Intel had a temporary competitive advantage because although MOS DRAM was first invented by
Intel which was valuable in the beginning but this was not rare and could be imitable. Mass
marketability of DRAMs brought larger competitors in the market by mid 1970s and Intel being a
small firm was not competing against multi-billion dollars Japanese conglomerates mostly into
semiconductors, consumer electronics, computers and telecommunications.
The difference between Intel’s and Japanese’s approach were as follows:

Intel Japanese
Intel had just started and was a small company Huge conglomerates
Capital was funded greatly by self Had access to abundant and cheap capital
through bank loans
Had access to little outdated technology Had access to advance technology
Produced DRAMs at relatively high cost and low Produced DRAMs at lower cost and higher yield
yield (50-60%) (70-80%)
Had a lower supplier advantage which reduced Had a higher supplier advantage which
yield increased yield
Intel was full generation behind Japanese firms Japanese firms were ahead in DRAM’s
generations

Following lead to Intel’s decline in DRAMs –

1. Mass marketing of 1K DRAM opened gates to the other competitors to strive for
price/performance.
2. Intel fell behind due to poor design and high cost as compared to competitors.
3. Intel was not able to catch up to the learning curve.
4. Intel was a generation gap behind Japanese when they introduced 64K DRAMs at mere $5-
$6 while Intel was still working on 16k DRAM.
5. By the time, Intel came with 1-megabit DRAM technology, relevance of DRAM was lost.
6. Intel also tried its hand in EPROMs which was read only memory devices where data could
be easily erased with ultraviolet and re burned again. Intel positioned EPROMs in the
marketplace in a very methodical and quiet approach with high price, thereby making it
valuable and rare but it gave them only temporary competitive advantage. After Japanese
and other competitors came to know about it, the market was diffused and Japanese came
up with a similar product with drop in prices by 90%.
7. Intel also invented and launched flash memory with higher prices but flash had much lower
volume and hence was limited to only four major players in the industry
8. Mass market of DRAMs and EPROMs was majorly captured by Japanese and Intel was
restricted to only 1% market share.
9. To gain 10% market share in DRAM business, there had to be a huge investment of $600
million in new fab but as industry was overcapacity, Intel chose to exit the DRAM business.

Why has Intel been more successful in microprocessors?

Intel invented the microprocessors in 1971 but did not realise its application till a decade unless
personal computers came into the picture. By that time, Intel’s microprocessors were used in
calculators and as components in industry controls. However, Motorola’s chip got its breakthrough
with the Apple’s first personal computer and that took away Intel’s first mover advantage.

Following were the moves made by Intel to put it back into the game:

1. Intel developed 8086 and 8088 microprocessors and initiated Project Crush in 1980 to boast
2500 wins through marketing and sales channel to secure 8088 micro-processor into IBM’s
first PC. PC market had huge growth potential owing to availability of packaged software and
less fragmented market and IBM had established itself as a highly successful PC division.
 However, IBM had a high buyer power over Intel and thereby forced Intel to supply its
micro-processor to dozens of companies to ensure adequate supply, also known as second
sourcing.

2. Project Checkmate was an intense marketing campaign by Intel which enabled large number
of design wins for the second-generation product of Intel – 80286 and reinforced Intel’s
position as the Industry standard, thereby giving Intel enough power to limit its licenses to
four companies.

3. Overseeing the disadvantages of second sourcing like giving away the technology to other
companies like AMD without a fair value and losing profits, Intel decided to be the sole-
source manufacturer and considered opening another plant in New Mexico instead of
second sourcing to AMD. Intel wanted to be the sole producers of CPUs so that it had full
control on the pricing.

4. Intel developed and launched 80386 which became extremely successful when Compac
pioneered the first 386 PC leaving IBM behind in the game and had set a new performance
standard. Consumers were willing to pay a premium for the additional performance.

5. Unlike with DRAMs, Intel was way ahead in competition in intellectual property like trade
secrets in design and production and patents which gave them substantial negotiating
power in cross licensing agreements. Creating patents and protecting their trade secrets
from AMD and rest of the second sourcing partners gave Intel the edge of differentiating in
the market, hence making its product valuable, rare and difficult to imitate. This offered
them sustained competitive advantage.

6. With the advent of RISC architecture which used fewer codes and lesser number of
transistors positioning itself as high priced, high performance in a niche market, Intel faced a
posing threat as all its microprocessors used CISC architecture. As it was not easy to replace
CISC with RISC in its existing software available in huge volume, there was no immediate
threat. However, Intel planned to launched CISC based 486 processor along with its new
RISC based i860. However, to subside the RISC based threat, Intel chose to back 486 and
positioned i860 as a niche product for specialized users.

7. Intel’s primary customers were OEMs of finished computers which sold PCs assembled from
five widely available standard components: microprocessor, motherboards, memory,
peripherals and operating systems and microprocessor accounting to 10-20% of the total PC
manufacturing cost. However, Intel had a complex relationship with their customers due to
the following reasons:

a. Intel used to advertise directly to the end users,

b. Intel vertically integrated into the design and manufacture of subsystems and full
systems
c. Intel was a sole source and it allocated supplies of newly launched chips in the face
of tight capacity constraints

All of the above posed high switching costs for the customers, giving Intel a higher power.
 8. Intel launched a new campaign called as “Red X” which featured a big 286in the full page ads
covered by a spray-painted red X. 286 was largely being sold by AMD at that time. The ad
targeted consumers to rethink on buying 286 when 386SX is offering higher performance at
same price. This campaign boosted the sales of 386 based machines.

9. Intel launch another campaign called as Intel Inside which involved co-op fund to reimburse
OEM’s half of the advertising costs at the expense of using “Intel Inside” logo in their
advertisements and stickers on their products. Intel faced internal strategic dissonance but it
was later covered with “Intel Inside” wide adoption by OEMs.

10. Intel also forward integrated by producing desktop computers in 1987 using 80386 without
screen, keyboard or brand. OEMs could rebrand them and sell in their own name. It also
invested in manufacturing a line of supercomputers that used 128 microprocessors.

11. Intel came up with the strategy to ramp up production of the new processors and allocate
only to the best customers thereby giving Intel more power than the buyer.

12. Intel also had high buyer’s advantage with the suppliers by posing itself as a sole investor,
thereby securing its privileged access to leading- edge equipment. Intel also formed a
consortium that included major chip makers to fund the technical work and to reach the
critical mass to gain industry acceptance and support for the technology. It also had dual
supplies to secure quality and timeliness of the equipment.

13. Intel and Microsoft had a symbiotic relationship as Intel was working on hardware platform
and Microsoft was working on software. A complementary software was critical for driving
demand of the hardware it produced. However, there was a conflict between them as
software production lagged behind hardware design. Therefore, Intel aggressively courted
independent software vendors to write applications for their leading edge processors which
enabled them to launch their product through various other programs like games or other
software applications which demonstrated higher performance using the processors.

14. Intel came up with its own “Intel Capital”, which invested in companies that fit strategically
into Intel’s business strategy and offered financial return. This was a cash cow for Intel.

Evaluate Intel’s shift in strategy under Barrett and Otellini?

Intel, which started with memory chip and then microprocessor anticipated need for new markets
outside PCs under the leadership of Barrett who became CEO in 1998 succeeding Grove. Barrett
spent $12 billion roughly on acquisitions and ventures during first three years as CEO pushing Intel
into new markets such as networks, communications, wireless and online services. He also changed
the mission statement and demonstrated Intel to being the preeminent building block sup-plier to
the world Internet economy. Consistent with Intel’s mission, Barrett reorganized company into four
areas : client platforms, server platforms, cellular and wireless and networking. Intel spend almost
$15 billion on capital additions and $12 billion in R&D in 2002-2004.
 - In Client platforms, Intel launched aggressive campaigns to drive one generation ahead of
AMD on all benchmarks in processors. However, AMD chips were largely found in the price-
sensitive consumer segment and this lead to the fall in the average prices and challenged
Intel’s historical revenue. Barrett pressed on raising performance at a reduced ASP to force
competition to innovate. There was again a strategic dissonance in Intel specially in the
development of mobile processor as the Vice President pressed on changing the
architecture of mobile processors along with performance and process.
- In Server Platforms, Intel introduced Xeon that served high end server segment that ran
corporate data centers. While servers was a small portion of Intel’s business, still 50% of the
Intel’s development went into servers. It came up with 64 bit Itanium processor. But it was
not gaining any advantage in sale due to its architecture and same was proposed to be
rewritten. Intel came up with 32 bit Celeron for lower end consumer PC market, 32 bit
Pentium for the mass PC market and 32 bit Xeon for the low and mid range corporate
servers and workstations. However, AMD’s strategy was to offer customers improved 32 bit
performance with some or all advantages of 64 bit CPU thereby giving AMD better
competitive advantage.
- Similarly in Cellular and Wireless, Intel came with XScale architecture which became the
industry leader in the performance at the low power range. Also, Intel’s PCA enjoyed wide
industry support with companies like Samsung , ATT developing products optimized for PCA.
- Communications and Networking industry was facing a recession and Intel’s main challenge
was to keep themselves ready to be the technical supplier of internet whenever it is ready.

Thereby, Barrett’s period saw lots of implementation errors like:

1. Loss of customers to AMD

2. Product recalls,
3. Product development delays
4. Cancellation of announced products
5. Technical miscalculations
6. Diversification
7. Retrenching
8. Failed ventures

The main cause of these failures was due to spread of Intel’s resources too thinly between the
maintenance between lucrative microprocessor business and ambitious pursuit of new business
developments. There was never a fixed goal.

Otellini was the new COO appointed to be in charge of Intel’s clients, server and communication
business thereby allowing Barrett to focus primarily on Intel’s new business areas and return Intel to
double digit growth. Otellini pressed on creating communications group indistinguishable from the
PC group.