Craig Barrett: How to cultivate an environment for growth in photonics
Craig Barrett is a leading advocate for improving education in the U.S. and around the world. He is also a vocal spokesman for the value technology can provide in raising social and economic standards globally. He was Chairman of the Board of Intel Corporation from May 2005 to May 2009.
Barrett has served on various committees to recommend policy directions to the U.S. government. Although not a member of the most recent committee, he is participating in the launch of the 2012 National Academies report, "Optics and Photonics, Essential Technologies for Our Nation." He was interviewed for SPIE Newsroom prior to the Washington, D.C. launch in September 2012 by Eugene Arthurs, CEO of SPIE.
Arthurs: You've been involved and I've seen your frustration with the lack of effectiveness of earlier efforts such as the two "Rising Above the Gathering Storm" reports. Do you have any recommendations, or hope, to offer that a report like Optics and Photonics can in fact have an impact?
Barrett: Well, of course there is hope for policy changes, and attitude changes and funding changes. Perhaps if you put this report into the framework of the recently released competitiveness survey by the World Economic Forum showing the United States sliding down positions in world competitiveness that people will eventually realize that political platforms should have more than statements about gay marriage and what the capital of Israel is and should focus on what makes the US competitive and what leads to economic growth.
Arthurs: You have a particular focus on the educational system, and certainly the Gathering Storm had very strong recommendations there. Are you hopeful about changing the educational system at the k-12 level?
Barrett: Well, there are some interesting changes in the K-12 system. One is certainly the increased concept of competition in education and that means competition in the form of charter schools or alternatives the standard public K-12 system. Actually, Washington D.C. is a leader in that area, with some 35-40 percent of the kids in alternative education systems. That trend has spread to about 40 of the states that now have some form of charter schools or competition. I think competition is always good to get the status quo changed, that's one positive.
The second positive is the movement to the state-driven Common Core Curriculum, which is basically upgrading the curriculum across 46 or 47 of the states in language arts or English, mathematics, and soon, science. And as the Common Core Curriculum gets introduced, that should be an upgrade to the curriculum. Then there's a follow-on to that -- a common assessment tool driven by the states to measure math, English, or science proficiency in an internationally benchmarked fashion.
We're making some progress with competition, we're making some progress with the curriculum. There's interesting progress I think with various universities and what they're doing with their schools of education and the pre-service education training really following the seminal work done at Texas with their UTeach program where the focus is now more on content expertise than the pedagogy of teaching. So, we're working to working to get more content experts in the classroom as opposed to pedagogical experts who know how to color between the lines in the right fashion. But other than that, we're getting to the point where you can start to put science teachers in science classes who know science, math teachers who know math, English, history, economics teachers who know their subjects.
So, there's some movement here. Of course, while we're making our movements in this area, other nations around the earth are also recognizing the importance of education and their upgrading their systems as well, so we have a long way to go to catch up.
Arthurs: I agree with you that the state initiative and the Common Core Curriculum is really good to see. SPIE certainly supports that, and hopefully we'll make some change. Let me go back to the photonics report for a moment. One of the recommendations is a national initiative, and I know that during the consideration of the report, there was a lot of talk about Sematech as a model. Is it a good model to look at for a national photonics initiative?
Barrett: I think there's a certain parallel aspect to it. Sematech was really created not so much as a basic R&D or basic technology initiative as it was a joint initiative among industry members to promote manufacturing capabilities. As the US was having little problem at that point in time from a basic technology standpoint, it was having a lot of trouble from a manufacturing/competitiveness standpoint. So Sematech was formed to really focus on the pre-competitive manufacturing capability in the United States. I think history has shown it did a pretty good job although there has obviously been continued competition from China, Korea, and Japan in that area.
My personal perspective is the key thing for either semiconductors or optics and photonics, or new materials, or biotech, whatever the topic you pick is the basic R&D funding provided by the federal government through agencies such as NSF, Department of Energy, and others. That's what really drives the concept of new ideas, new technological breakthroughs coming forward. Corporations have an important role to play in their research activities, but in the US their research activities are generally more in the applied, product-driven nature.
I think the key issue here is really singling out photonics and optics as one of the key technologies of the 21st century. And by the way, just about every other major economic power has done the same thing already. Whether it's Europe or China or Japan, or you pick the economic power -- they've all listed optics and photonics as a key area. And unless the US continues to aggressively pursue research and basic technology development in this area, we'll fall behind. I think that role of the government to fund basic research in our national labs, and our number one research universities is probably the most important thing that we can accomplish.
Arthurs: I know of Intel's interest in silicon photonics as we call it, and there's been a fair amount of R&D done at the universities. Yet, if somebody at a university wants to actually produce product from his modeling or research, it is extraordinarily difficult and right now I think they would either end up going to imec in Belgium or to Singapore to get prototyping done. Intel, I know, has really been pushing, at our meetings certainly, for some national capability for doing that. Do you see silicon photonics to address the bandwidth and energy issues for processors as important in going forward?
Barrett: I think there's no question that it's important in going forward. The roadmap in the past has been pretty much Moore's law on the one hand. There's a series of laws that kind of go hand-in-hand, Moore's law from a transistor standpoint. There's been a memory density law from a hard drive standpoint. There's been a photonics law for how many photons can push down a glass fiber. There's Metcalfe's law for the value of networks like the Internet going as the square of the number of nodes. Those three or four laws have been what has driven the explosion of the Internet over the last decade or two, and to continue to move the whole Internet society forward, you need those laws to continue.
Silicon photonics is one great step in that direction to help the combination of Moore's law of communication, bottleneck, etc. not only at the local level, at the chip level but as you expand to the server level, to the data center level as you look at connecting different data centers with huge bandwidth pipes. So I'm aggressive about what Mario [Paniccia] and his team at Intel have been doing for the last several years. I think they've made great advances. It's always a challenge though with the big capital costs associated with facilities that operate at the state of the art, much as it is with simple semiconductor devices. You don't find many universities or national labs with the capability that Intel has or a state-of-the-art semiconductor supplier has.
There are probably lots of options. You could talk about having a national facility that's expensive that would be nice to have. You could talk about arrangements with the private sector for capability for universities to work with them. That's also a possibility, I think there's lots of options in that space. But, it is clear that you can't move the technology forward very well without the ability to imprint the technology in actual materials and make devices and study how they operate.
Arthurs: There's been a push to actually get some silicon photonics capability prototyping in the US, such as at OpSIS, and hopefully that will continue to succeed. I'm hearing from you, and don't want to misinterpret, that we really as a society should be pushing the agencies to continue to increase the funding at the fundamental level for optics and photonics.
Barrett: If you went back to the National Academies reports, the Rising Above the Gathering Storm reports, that was absolutely a key recommendation made in 2005 and then 2010. That wasn't the first time that recommendation has been made. The issue had been festering in the technical community well over a decade before that. I think that the biggest opportunity that the US still has is to take its investment in basic technology research back to same level as a percent of GDP as it was 30 or 40 years ago. And it's down substantially since that time frame in terms of percent of GDP. As you move into the 21st century and you say it's the technology century, the innovation century, the knowledge century, you can't expect to be very competitive in that century if you're continuing to lower your relative investment in those technologies. And that is precisely what the US has been doing.
I said in my earlier comments that I would have been much happier watching the Republican and Democratic national conventions if they had just taken Rising Above the Gathering Storm as their platform as opposed to what they did put out.
Arthurs: Certainly I think that was widely accepted across Congress and they didn't appropriate the funds for it.
Barrett: And that's happened before in my memory. Everybody has their theory on that, and probably the most common theory is that basic R&D pays back in a 5-10 year time frame. Elections are in a 2-year time cycle and one doesn't necessarily support the other.
Arthurs: You mentioned in your "exit interview" from Intel that the tax structure needed to be addressed and I certainly heard Norm Augustine speak on that quite a lot. Do you think there is any hope of that taking place, the incentives basically being changed for long-term views for corporations rather than the current tax structure which incentivizes everybody for the quarterly results?
Barrett: I guess the good news on the corporate tax issue is that at least it's being publicly debated today, and there are calls for effectively trying to simplify the tax structure but also to make it more competitive on an international basis such that the tax structure is not a disincentive to invest in the US. You hear lots of politicians talk about how corporations should be patriotic and invest in the US and they shouldn't hire any employees outside the US. Both of those statements, from a private perspective, they're kind of naïve. If you have great technology companies like IBM or Intel or Boeing or Microsoft or whoever, who are doing the bulk of their business outside of the US -- well over 50 percent -- the concept of not having foreign employees is very naïve. The concept that companies should be patriotic as opposed to serving their shareholders is also a bit naïve.
So you have to in fact provide a local economic infrastructure and a local economic environment that promotes investment in the US and not disincentivize investment in the US. I personally took a very simple one-page summary of the company that I used to work for, which was going to build a multi-billion-dollar facility and looked at the economic structure in a country like the US with a high tax rate compared to countries with a low corporate tax rate, and the net present value difference was well over a billion dollars and it was driven primarily by tax and incentive issues, not labor issues. So, if the CEO of Intel were to make a decision to build a 4- to 5-billion-dollar manufacturing plant in the US compared to a lower tax environment somewhere else in the world, they would be making a decision to spend an extra billion dollars over that plant's lifetime. You don't have to have an MBA from Harvard to understand the disincentive nature of the tax rate.
Arthurs: The countries that have identified optics and photonics as important, including the EU, not a country but an economy and Germany being a prime example, have really taken a top-down approach and invested very heavily and "put their finger on the scale." When that concept is mentioned in Washington, it tends to run into a belief that the market will decide and it's not government's business to do any of this.
Barrett: I think there are various ways in which a national government can get involved. One way is in fact to do a top-down, winners-and-losers pick, create a facility and do it in a national-lab type environment. Another way is to basically let the marketplace decide but make the decision that you're going to invest money in that area and then let the marketplace decide by supporting peer-based research activity and let the best ideas win as opposed to letting a government bureaucrat direct the program.
I think the history of the US in supporting research activities has been peer-based evaluation of research ideas and then the federal government funds the best idea -- not the best idea decided by the government but by experts in the field -- and then let the best person win on that basis. I think that's the power of the US approach. Occasionally we do things like let's go to the moon, for ideas and major big programs. Those big mission-directed programs are usually not so much state-of-the-art technology development as they are utilization of existing technology to accomplish a mission. What we are really talking about here is the creation of new technology, not applying existing technology to satisfy some desire.
Arthurs: I hear exactly what you're saying. I do think that our peer-reviewed scientific enterprise, though, has become increasingly separated from the economy and from manufacturing of the US.
Barrett: That probably depends where you're looking. My impression is that the great strength of the Tier 1 research universities in the US, you pick the one of your liking, is that they are stacked with incredibly bright faculty and students, they generally have reasonably close association with industry, they're not void of industry association through either consulting or exchanging students or professors or what have you, and that the research they do is quite often very supportive of what industry has going on. The best measure I have of that one, by the way, is when I visit the European universities and see the direction that they are moving in, it is in fact to copy exactly what I've been talking about, the Tier 1 US research universities.
If you look at what the Russians just did with their Skolkovo project, it is to contract with MIT to transfer MIT DNA, the infrastructure, the way that MIT is run to create a Russian university of exactly the same character. Around the world what I see are universities moving to the US Tier 1 research university model, which I think, frankly, is the great backbone strength of the US technology infrastructure.