Intel Labs Case-Study

Intel Labs (A): Photolithography in crisis - Case Analysis

·         Intel is a very good example of an ‘Ambidextrous organization’.

·         Their operational divisions are very mechanistic with very structured hierarchy, repetitive processes and documented standards.

·         Their Research & Development has been more Organic and has morphed and evolved based on the demands of the business and the external environment.

·         They currently have a very globally linked Strategy for R&D.

·         Some of the limitations that Intel is observing currently are the economies of scale and learning curve effects. Funding for projects has become contentious and every research division is working on projects based on its own market and requirements.

·         They also enjoy some advantages of their strategy. They have a huge network of research resources from various sectors of the industry and also from outside of the industry.

·         Their organic structure allows them to be fluid and adapt to the requirement of a project.

Recommendations:

·         Wilson should take advantage of Intel’s organic and decentralized research organization.

·         Garner the assistance of the technical fellows of Intel to further analyze the possible alternatives and any concurrence with current Intel R&D efforts.

·         He should identify the resources from within Intel’s network that can be deployed for the development of the EUV photolithography technology.

·         Combine the access to the resources that will be available due to lack of government funding and the resources within Intel network.

·         Leverage the organic nature of their R&D department and form a project team that can channel efforts into development of breakthrough technology.

·         Utilize their relationships with vendor organizations to fund their research that helps build complementary technology. (Boeing is an example. Though Boeing had its share of troubles, the business model is already a basis for Intel’s Fabs and its R&D efforts so they are more familiar with the concept of outsourcing)

·         This will allow the EUV photolithography to become the dominant design of future and paves path to commercializing the research and benefitting from their strategic investment.

_______________________________________________________________________________

Chapter 10 and Intel Labs:

Firm Size and Impact:

Large Firms- Positive Aspects

• Economies of Scale in R&D
• Greater access to complementary resources
• Learning Benefits

Large Firms- Drawbacks

• Governance problems
• Not as agile
• Less centralized control

Larger companies can be broken up into smaller divisions, which can feel like small companies within a large company.

Other effects on Innovation within a firm:

Standardization-improves efficiency, roles and responsibilities are performed in a unified manner

Formalization- improves efficiency, utilizes rules and procedures

Centralization- Decision making authority is kept at high levels of management

Decentralization- Decision making authority is apparent throughout many levels of the organization

Two types of Organizational structures:

• Mechanistic- highly formalized and standardized, ideal for efficient production
• Organic – free-flowing, ideal for creativity and experimentation

Companies can be ambidextrous and benefit from both efficiency and creativity/experimentation; they do this by alternating the structure of the divisions within their company.

______________________________________________________________________________

Intel Labs:

Basic Research and R&D:

One thing that stood out to me in this case is Intel’s approach to research. They seem to have their organization built and their business model dependent on “minimum information.” This causes the ideas to be limited and the basis for exploration and discovery to be stifled. I think Intel should consider including a research component to their organization, it may generate more ideas than necessary and Intel may not be interested in academic pursuits, but it will also help them be cutting edge, industry savvy, and contribute to a larger industry presence and awareness.  Basic research would also allow for more responsive; something that the “Copy Exactly” model doesn’t prepare them for, as stated in the case, it is a good model for incremental innovations but not for larger discontinuous changes. Decentralized R&D labs also made for disparate project work for Intel, there was little communication and cooperation across groups.

What are the options in the case?

1) Utilize connections at universities across the US and try to establish a research program for EUV.

2) Abandon EUV and let other technologies around lithography emerge before acting

3) Change Intel’s philosophy and bring research and commercialization in-house by hiring researchers

4) Find investors outside of Intel to fund the development of EUV

What should In Intel do?

I think Sandy Wilson should consider options 3. Abandoning the project after so many resources and so research has been put into it wouldn’t be wise. They have been unsuccessful at raising funds thus far, so that doesn’t seem to be a viable option either. A major missing component in Intel’s R&D structure is basic research in-house, as I mentioned before. So, I think they should make a change to their organizational model and consider developing and commercializing the product within the company. They can benefit from the research, from the marketing attention, and ultimately with all the potential EUV shows they can potentially benefit from providing a valid new light source for lithographic technology.

_______________________________________________________________________________

Tariq Haniff:

Summary of Readings:

            In this week’s reading we explored how a structure should be organized to maximize its innovation and throughput of Research and Development (R&D). We begin by discussing what impact the size of a firm has on its ability to proliferate R&D and innovation. Arguments are made regarding the impact that the size of a firm has on its ability to innovate. We take  a look at the strategy of disaggregation and modularity.

            The “structural” organization of a firm can also have a great impact on its ability to create and innovate. Formalization and Standardization describe the degree to which a company governs its employees, procedures and policies. A high degree of standardization and formalization , and the associated stringency, are seen to be less conducive to innovation. Corporations struggle to balance mechanistic and organic structures to ensure they can operate in an efficient manner regarding production and output, while still nurturing an environment that encourages free thought and creativity. In doing so, they strive to become ambidextrous organizations.

            Corporations have different options regarding the management of their R&D and innovation. Building “modularity” of flexibility into a design allows a corporation to standardize its current products to meet many different customer needs while focusing its R&D on new products. Some corporations choose to centralize their R&D efforts, creating an insular and autonomous research center, while others form a “loosely coupled organization, consisting of many small R&D centers that have a high degree of autonomy, yet leverage recent improvement in communications and IT to integrate and coordinate their efforts.

            Finally, the management of innovation across borders is discussed. Many multinational corporations are face with this challenge. Two strategies: “Center for Global” and “Local for Local” (as their names imply) are presented.

Case Questions:

1.      What lessons did Intel learn from the days of the founders at Fairchild?

Namely that collaboration and integration between production and R&D is essential for a technology company’s success. They initially succeeded by requiring all new researchers to become intimate with the production facility capabilities.

2.      How should Intel approach the development of its next generation lithography?

It seems that the EUV technology is closest to the standard UV lithography currently in use by most semiconductor companies. The X-ray lithography approach looks to be capital intensive and disruptive. A research consortium with several other companies would be my suggestion since interest from the government and National labs has faded.

3.      What risks is Intel taking by not being fully active in the development of EUV lithography?

If Intel allows the tool to be developed independently, they would be at the mercy of the equipment vendors. Not participating would give competitors the cooperative advantage.

Chance Durham

Chance Durham

GMAN 509, Cohort 15, SR

Professor Dr. Jyoti Bachani

Session 7 -Chapter 10: Organizing for Innovation

Intel Labs (A): Photolithography Strategy in Crisis (HBS)
<!--[if !supportLineBreakNewLine]-->
<!--[endif]-->

Intel is faced with an innovation strategy crisis because of a shift in the government’s willingness to develop technology to keep up with the expected rate of semi-conductor advancement (Moore’s Law).  Intel has become a consumer of chip manufacturing technology and an innovator of chip design and manufacturing process.  Three emerging manufacturing technologies are on the horizon that has the potential to manufacture ever smaller nanometer sized semi-conductors.  The technologies include Xray Lithography XLR, ION Projection Beam, and Extreme UV lithography.

 Intel’s distributed model of R&D in which each lab using the exact same equipment for development in order to leverage “Copy Exactly” methodology is expensive to maintain when technology takes on radical change, but works well with incremental change.  Of the 3 potential technologies to advance the CPU design complexity through increased miniaturization, the Extreme UV is the least radical change, can leverage more of the existing equipment, and has the best potential to advance the circuit line size down to .03 micron (beating the closest competing technology Ion Beam Projection at .10 micron).

What should Intel do to not let the Extreme UV technology perish with the government spending cuts at Livermore and Sandia National Labs?

Intel could cut a deal with the government to fund the further development of the Extreme UV technology by offering to pay the salaries scientists working on the project in exchange for the continued use of Livermore and Sandia Labs facilities.  This relationship could give Intel better access to develop the technology with manufacturing in mind, and to standardize the laboratory to their fabrication facility.  This would be least disruptive to the progress being made with Extreme UV, and allow both the government and Intel to first access to <.03 micron semiconductor technology.  The increased computer power made available through the semi-conductor innovations certainly would have military application, and Intel could maintain first mover advantage into the next generation semi-conductor design and manufacture. 

How can Intel maintain a competitive advantage while sharing the cost and risk of developing the Extreme UV technology?

Intel could make an acquisition of the Extreme UV technology from the government and then use its relationship with universities to fund and build Intel standardized labs, at the schools, used for the developmental completion of the technology.  Then the “Copy Exactly” method can be used to move the Extreme UV into Intel’s own production fabrication facilities of the same design as the labs at the universities.

What is the risk of Intel’s reliance on other parties to develop the needed innovations to move its semi-conductor manufacturing capability forward?

Although the cost of developing next generation semi-conductor manufacturing technology is less by letting other entities, such as government labs or IBM, do the R&D, the risk is that Intel’s ability to remain competitive is in the hands of outside parties.  Historically, savings on R&D has worked out to Intel’s advantage due to a focus on manufacturing efficiency which has led to the ability leverage emerging technology in such a timely manner that competitors have not gained a first mover advantage.  Without an adequate stake in the R&D of next generation chip technology, Intel runs a risk of being shut out of a radical technology innovation,  not being able to adapt fast enough, and then succumbing to the first mover advantage and dominant new design of a competitor.

_____________________________________
Diane Reyes

Chapter 10: Organizing for Innovation

Intel Labs: Photolithography Strategy in Crisis

            Currently, Intel is run as a mechanistic organization.  Their approach to manufacturing and setting up their fabs trough different parts of the United States is called Now Copy.  In this mechanistic approach for example, in 1989 their fab facility in Albuquerque was designed to receive the i486 from the development lab in Oregon.  This included the same equipment, the same options, the same physical dimensions of the plant, and even specifying hook ups of components within their equipment.    Through Now Copy, Intel was using the same configuration options on the same equipment from the same vendor.  This newly defined process helped Intel capitalize on the very expensive equipment they were using, but a limitation was that this process did not work well with discontinuous products.  Rather, it was designed for incremental changes within a product.  In 1996, Intel spent two thrids of its $1.8 billion in R&D in new product development and one third in process improvements.

            Intel’s competitors IBM and AT&T had a different approach to R&D leveraging partner companies to maintain large efforts on basic research.  The competitors’ partners had research labs in other states and even with other companies located in other states.  There was an advantage in this strategy since IBM was able to get first dibs on equipment for emerging technologies and products which helped them develop more advanced products and led to a greater competitor’s advantage.
            With Intel’s strategy of pushing to prioritize manufacturing then research, it leaves Intel executives in quite the dilemma of LLNL losing research funding from the federal government with all research that do not have direct military applications.  The option that I would recommend that Sandy take is to take advantage of Intel’s national research connections and try to assemble a research team from there.  This involves getting research students and faculty to work together.  This would be the cheaper option.  By leveraging knowledge from universities, Intel can hone in on creative ideas that have no bias form past Intel research.  The incentive for students for their graduate research is there and driving this factor is a benefit for Intel. 

            Another option that can also work but is the opposite of this option because it is more costly and it involves leveraging researches that have experience in the EUV technology is for Intel to fork out dollars for EUV research.  This meant that Intel would have to utilize their current researchers as well as unemployed engineers and scientists from those who lost their jobs due to the federal funding cuts.  Costs will need to be spent on new research facilities, new equipment, and a new approach to research which will be out of Intel’s comfort zone since they have mastered their Now Copy approach.  High risks yield to high rewards, however Intel will need to determine their risks with this options as their competitors have already been practicing this method and may be ahead of them.
--------------------------------------------
 Chris Hilgers

Chapter 10:

·         No framework provided for deciding on the most appropriate structure given firm need.

o   Put forth all of the options and benefits and limits of each, but no recommendations made

§  Cited examples of companies that successfully implemented strategies, and instances where they were not beneficial

·         Definitions only chapter?

·         Would have liked more firm examples for innovations across boarders

o   Innovation driven by cultural reasons that was then available to world?

·         Clearly there is not a one size fits all model, but what evaluation criteria should be applied?

o   Do you simply introduce one arbitrarily and test out each system on an as needed basis?

·         Google case was interesting in that they have most of their divisions follow a “normal” structure, but only have the “chaotic” organization in specific sub-units

·         How is my company organized? 

o   Centralized – key decisions made at top levels

o   Can a company have predominant features with “one off” scenarios? Even more so than Google.  If one company unit has autonomy while others do not, does this breed jealousy and impact working relationships?  Need more of a focus on human element impact.  Obviously, someone will be upset if you change what they are doing (3M case), but what if you offer “perks” to another unit not available to their unit?

Critical Case Questions: (use evidence from case, be specific!!! Don’t give generic statements that work for all companies)

Intel Labs (A)

Goal: to fund the project with national labs for EUV lithography in the next five months to “get a seat at the table”.

While I do not have the necessary technical background, the case did not seem to express why the pursuit of the EUV lithography was so necessary/important.  I understand that limitations outlined for the technology after the next five years, but why did Sandy believe this showed so much opportunity?

1.      Option: Intel connections at universities to assemble research program.

·         Would disaggregate current research process in place at Livermore Labs

·         Not clear if able to keep current research team in different locations

·         Would be difficult to maintain “Intel” formalization/standardization, would be dependent on Universities methodology

o   Not inconsistent with current framework from Livermore Labs

·         Would Sandy be able to centralize decision making authority around his position?

o   What if different universities pursued different incongruous areas of research

o   Or followed information that led them away from EUV commerciality

·         Could prove to follow successful loosely coupled organization structure

2.      Option: Do nothing, wait for other lithography technology to emerge.

Doing nothing would ultimately end the EUV project.  Sandy just took on this new position, could he possibly recommend that Intel not pursue it?  Wouldn’t it mean the end of his position?  Is this a legitimate option Sandy could have pursued?  Is it a legitimate option he should pursue?

3.      Option: Reverse Intel philosophy, invest the money necessary to commercialize EUV

·         Could replicate current research process in place at Livermore Labs

·         Could better leverage assets of Intel, more mechanistic structure

·         Could be implement “Intel” formalization/standardization

o   Inconsistent with current framework from Intel

§  Untested

§  Could upset current R&D, jealousy

·         Ambidextrous Organization

o   Is Sandy trying to get the best of both worlds

o   Would this look more like a power play where Sandy is trying to get more autonomy and power

Recommend Option 3 w/concerns above

----------------------------------------------------------------------------------------

Jason Au:

Chapter 10 Notes

* Google Case– I found Schmidt’s comment regarding their “chaotic” organizational structure interesting.  Although creativity is aggressively promoted in certain departments within Google, many departments are still ran in a very traditional way.  This shows that an organization can have multiple facets to its identity and company culture.

* Advantages & Disadvantages of a Company’s Larger Size

                -Pros: better financing options, economies of scale, learning effects   

                -Challenges: loss of managerial control, less nimble/responsive to change

* Formalization and standardization can lead to increased productivity and efficiency, but could also inhibit innovation.  Centralization can possibly enhance or impede flexibility and responsiveness.

* 2 Types of Structures:  Mechanistic (highly formalized, standardized) – yields greater operational efficiency ; Organic (free-flowing) – better for innovation and dynamic environments and yields more creativity

Intel Labs Case

Key Notes:

·         Photolithography expected to have a shelf life of 5 more years

·         EUV developed by Lawrence Livermore National Labs and Sandia National Labs

·         Intel already invested several millions dollars to track this research

·         Intel had a production focus and steered away from having a dedicated development fab.  They created "Copy Exactly" as a method of having the same machinery and processes across the organization

·         Per Gargini, “Copy Exactly” was good for incremental innovation.  Doesn’t work well for large discontinuous changes in technology

·         1/3 of R&D went towards process improvement, 2/3 went to new product development.  Intel’s spent a total of 1.8 billion in 96

·         Other large semi conductors ran R&D differently and more traditionally and worked closely with suppliers...they got an advantage this way by getting new units early.

Intel’s Options:

(1) Leverage research relationships at the university level; (2) Abandon EUV; (3) Invest and commercialize EUV;  (4) Convince other companies to invest with Intel in EUV development

Recommendation:

Since Salvador has already made attempts to recruit other companies to step in to fund the research costs with no success, he should consider option (1) and leverage the existing relationships at the university level to help research and further develop EUV technology.  Intel has already spent a considerable amount of money following the labs research.  To date, they don’t have any other options to compete with their other competitors, IBM and Lucent’s Bell Labs.  Partnering with the universities to develop EUV will not only assist with “bringing them to the table” alongside their competitors, but it could also help them attract and recruit the future talent of the industry to keep Intel competitive.  Intel should also consider changing their fragmented R&D Copy Exactly structure to allow for more innovative and creative technology opportunities. 

---

George Fuentealba
Ch. 10 and Intel

Chapter 10 Insights and Key Learnings       

·      The structure and degree to which an organization uses formalized and standardized procedures and controls can greatly impact innovation. Additionally, the size of an organization’s divisions may or may not stifle innovation and success.

·      Firms that are divided into networks of smaller, often more specialized, autonomous divisions are referred to disaggregated (also termed as virtual, network or modular organizations)

·      Firm size: Bigger firms advantages include (1) better to obtain funding, (2) spread R&D costs, (3) may increase economies of scale; whereas bigger firm disadvantages include (1) R&D efficiency decreasing, (2) more bureaucratic “red” tape

·      3 structural dimensions presented in chapter are (1) Formalization refers to degree to which rules and processes structure behavior of employees; (2) Standardization refers to degree that activities are performed in a uniform matter; (3) Centralization: degree to which decisions are made at top or activities performed at central location.

·      Mechanistic structures have high formalization and standardization. Organic have low formalization and standardization (more “free flowing”).

·      Ambidextrous organizations: ability for organizations to behave as 2 different kinds of companies: different divisions with different structures and control systems while enabling diverse cultures and patterns in operations.

·      Google is a good example of an ambidextrous structure. Normal business operations are mechanistic whereas R&D are organic.

·      For multinational firms, Bartlett and Ghoshal identified 4 key strategies: (1) Center-for global: everything centralized. (2) Local-for-local: each division has own R&D that accesses local diverse resources. (3) Locally leveraged: each division does own R&D but attempts to leverage diverse ideas. (4) Globally-linked: decentralized R&D labs that plays different role in firm’s strategy.

Intel Case

In the Intel case, Sandy Wilson is troubled with what to do about Intel’s future investment in a newer extreme ultra-violet (EUV) light research and development that will be a game changer in the semiconductor industry. EUV technology will allow for shorter, non-visible wavelength light to create print lines of less than 0.10 micron width to pack more information in each wafer. The smaller the micron, the better the technology.

Historically, Intel had undergone several structural changes. Most recently, it has adopted a modular, mechanistic approach for how it’s R&D divisions were set up. Intel chose to build its organization around manufacturing and then the research using a “minimum information” principle. Intel created their own methodology early on for setting up new R&D facilities called “Copy Exactly”. Copy Exactly describes using the same configuration options on the same equipment from the same vendor at each facility.

By 1996, Intel determined it needed to organize its research activities more formally. They looked at it more internally and externally. Internally, Intel created a decentralized distribution model of research by creating 3 types of labs: Intel Architecture Lab (IAL), Microcomputer Research Lab (MRL), and the Components Research Lab (CRL). Externally, Intel would conduct research through funded university programs, external research channels and equity investment programs. 

In order to compete in the photolithography space, Intel had been working closely with the Lawrence Livermore National Lab on the EUV technology, originally funded by the government. EUV was far superior to the existing technologies of X-ray, Electron beam and Ion beam by their competitors. However, the government had decided to stop funding EUV and Intel’s EUV lithography projected was slated to end in five months. Wilson had outlined four options:

Option #1: utilize university research programs to help continue EUV research.

Option #2: Abandon EUV and wait for other lithography technologies to emerged amongst competitors

Option #3: Reverse Intel’s philosophy, and invest money to commercialize EUV inside Intel, requiring new facilities, equipment and approach to research (new labs).

Option #4: Develop new ways to others to invest (already tried and failed by predecessor).
My recommendation would be for Wilson and Intel to attempt option #1 and utilize existing university programs to conduct the research. Since they have undergone several structural changes over the past few years, their current decentralized structure is already working well for other product line. That said, innovative products lack a home in their current structure. By utilizing external programs established, they can reduce their internal investments in new equipment and facilities.