Reed's Ruminations: A Blog by Dan Reed

Consilience: Innovation Enabler

A few thoughts from my Communications of the ACM blog post.

In this era of hyper-specialization, it is important to remember the combined power of generality and specialization, lest we render true the old joke that education teaches one more and more about less and less until finally one knows everything about nothing and is eligible for a Ph.D. After all, Ph.D. is an abbreviation for the Latin Philosophiae Doctor (Doctor of Philosophy), and science itself evolved from natural philosophy.

I Could Smell the Smoke

Certain dates take on special meaning, as shared experiences and defining vignettes in our global consciousness. Everyone remembers where they were and what they were doing on these dates. The Apollo lunar landing (July 20, 1969) is one such date. For many in the United States, September 11, 2001 is another. The recent anniversary of this date prompted me to reflect on my experiences

On that day, I was hosting a National Science Foundation (NSF) workshop on Computation as a Tool for Discovery in Physics at NCSA's Access Center, adjacent to NSF, in Arlington, Virginia. After the meeting opened, I ran downstairs to hail a taxi and head to Capitol Hill for a meeting at U.S. Representative Dennis Hastert's office in one of the Congressional office buildings. At the time, Rep. Hastert was Speaker of the House and second in line to the U.S. Presidency, a fact I had not considered particularly noteworthy until that day.

I arrived a few minutes early for my appointment with one of Rep Hastert's legislative affairs directors and was waiting for my other Illinois colleague in the hallway outside when my Blackberry began buzzing. Argonne's Rick Stevens had sent me a message asking if I were okay. I thought this curious and responded, "Of course, why wouldn't I be?" as I ambled into the office anteroom. Like almost all Senate and House offices, the anteroom contained televisions tuned to CSPAN and CNN, with the sound off.

At that point, I heard a mumbled comment about a plane hitting the World Trade Center. My immediate thought was that the pilot of a single engine plane must have done something incredibly stupid. As I was thinking this, the person I was waiting to meet emerged and told me we would need to postpone our meeting. "Why?" I asked. Pointing to the television, which was now showing a replay of American Airlines Flight 11 hitting the first World Trade Center tower, he said, "That's why."

Unbeknownst to either of us, a third plane had just hit the Pentagon, a few miles away. As we stood chatting, the Capitol Police began a leisurely evacuation process for the building. As I emerged into the fall sunlight, small groups of people were wandering slowly around the House and Senate office buildings. There was no panic, because most of us had no idea what was happening, at least until we looked to the southwest, where a column of smoke was now rising from the Pentagon.

My Illinois colleague and I retreated to the Bullfeathers, a bar just off the Capitol, where we joined a growing group of members of Congress and staffers. Rumors were rampant, including one that a plane was headed for the Capitol, and none of us really knew what was happening. Cell phone service was non-existent, overloaded by call attempts, and we had only relayed information from the single television in the bar. Through it all, my Blackberry continued to operate, though, and I received brief news updates from my friends.

Why didn't we leave a high value target area, you might ask? The Washington Metro was closed, taxis were non-existent, there was a security cordon around the area, and it was too far to walk. As we waited, fighter jets were flying low over the Capitol, sirens wailed regularly and bomb squads flew past.

After several hours, some semblance of normality began returning and the Metro reopened. My colleague and I boarded the Metro, which was packed with worried people. Their fear rose further when the train passed the smoke-filled Pentagon stop at high speed.

After what seemed a lifetime, but which was really only about three hours, I found myself again sitting in my meeting at NSF, discussing the future of computational physics. All the attendees had agreed that continuing the meeting was the best policy. It was somber and subdued, and we struggled to focus on plasma physics, molecular dynamics, or accelerator physics.

We now knew the airports were closed, and most of us knew we would need to find alternative transportation to our varied homes across the country. At least one of the workshop attendees had immediate family within blocks of the New York ground zero site, which triggered collective attempts to find any working communication channel to New York City.

Throughout the day, I continued to track developments on MSNBC, where I began experiencing timeouts due to query volume. That evening, I walked through the deserted streets of Arlington to reach my nearby hotel. Seeing the National Guard armed and ready outside buildings was sobering. It was a surreal time, with great uncertainty about what might happen.

Over the next three days, I collected information on various faculty, staff and students stranded in greater Washington, DC. Eventually, we found a rental car (no mean feat) and drove back to Champaign, Illinois.

Impact, Not Indicators or Artifacts

N.B. I also write for the Communications of the ACM (CACM). The following essay recently appeared on the CACM blog.

Publish and/or perish; proposals and reports; research, teaching and service: these are the "death and taxes" equivalents for life in major research universities. Success -- or at least promotion and tenure – is normally measured by how well one plays the academic game, publishing frequently in the "right" places, placing students in the "right" institutions and serving on the "right" committees. All too often, however, we tend to focus on these indicators rather than the presence or absence of impact from our activities.

Counting Beans

Many of us know faculty who have published hundreds of papers, but who have not demonstrably changed the nature or direction of their specialty. Conversely, many of us can readily identify individuals who have transformed the nature of computing, but who have few – if any – of the standard beans that are counted when quantifying traditional research success.

As a department head, my guidance to young assistant professors was always simple: do something interesting and important, make a difference that is recognized, and the artifacts will come. The artifacts are just that -- things that accrue from impact, not objectives to be sought.

I was reflecting on all of this as I tracked the responses to Moshe Vardi's CACM editorial, Conferences vs. Journals in Computing Research, and Lance Fortnow's CACM article, Time for Computer Science to Grow Up. Before these two articles appeared, I had already discussed these same concerns about computing's shift from journal to conference publications in my March 2009 CRA column: Publishing Quarks: Considering Our Culture, where I bemoaned the rise of the minimum publishable unit and called for a reinvigoration of journals.

Cultural Conservatism

It is somewhat ironic that many aspects of our computing research culture are now so often conservative, especially when research itself is defined by an insatiable curiosity about the unknown and unexplained. If you have any doubt, try to secure research funding for a proposal that is controversial or try to move a curriculum revision rapidly through the academic approval process! (To be sure, there are exceptions and there are good reasons for many of these checks and balances, but the process is sometimes excruciatingly slow and often incremental.)

It is worth remembering that our field grew from the confluence of mathematics, physics, electrical engineering, psychology, information theory and management and a host of other disciplines. The founders were radicals and iconoclasts, eschewing the extant culture of their disciplines because this new notion of computing was so pregnant with possibilities that could transform society and the nature of discovery. In short, it was the most exciting thing they could imagine doing, even if it was not the "safe" thing to do. More to the point, they were not focused on getting multiple papers in this year's specialty conferences.

Remember, it's the impact that matters, not the artifacts.

Language Shapes Behavior: Our Poor Cousin, Data

The Sapir-Whorf Hypothesis holds that language and culture shape behavior and thought. The idea has a long, rich and sometimes controversial history, though recent cognitive linguistic research, together with brain imaging, has lent it new credence. Lately, I have been reflecting on the nature of linguistic relativity in scientific research, particularly for HPC and cloud computing. In particular, how has our focus on floating point operations per second (FLOPS) shaped discourse, science politics and outcomes?

Look, It's Terascale!

If I were to remark, "That is a terascale system," you would likely interpret my comment as an assessment of the system's computing performance, more specifically its floating point capabilities. For the cognoscenti, the sentence likely engenders visions of multicore systems with out of order instruction issue and deep pipelining, small clusters with fast interconnects and (perhaps) some accelerators such as GPUs. You might even reflect on the historical evolution of high-performance computing, when a terascale system was ranked first on the Top500 list about a decade ago.

Given my remark, you would be unlikely to eye your desktop computer with its terascale (terabyte) commodity disk drive. Yet, based on its storage capacity, that desktop is just as surely a terascale platform as is the compute cluster. Ah, you say, I have mixed capacity (bytes) with capability (operations). Terabyte per second storage systems are rare, you say. Absolutely true, yet as this bit of third rate linguistic legerdemain illustrates, language and, perhaps more importantly, connotation really does trump denotation in our discourse.

Whether Petascale and Exascale?

Today, the world's fastest computing platforms have broken the one petaflop mark on the high-performance Linpack (HPL) benchmark, and planning groups across the United States, Europe and Asia are debating political and technical approaches for the construction of exascale computing systems. I cannot help but wonder, though, why we in technical computing talk so little about petascale and exascale data platforms? Why is there no international initiative to build a low latency, high bandwidth (many terabytes/second to even a petabyte/second) data analysis engine with multiple exabytes of storage capacity?

I suspect the reason lies in our background and linguistic heritage. Most of us in high-performance computing came from mathematical backgrounds, where success was defined by a proof and an equation and their embodiment in a computational model, rather than by insight gleaned from large-scale data analysis. I believe it is time we found a lingua franca that bridges FLOPS and bytes and reclaim the full connotation of exascale.

I Want My Exabytes

It is worth remembering that the explosive growth of observational data is itself largely a product of inexpensive CMOS sensors, based on the same semiconductor technologies that begot microprocessors. The examples of high resolution instruments are legion, from the Large Hadron Collider (LHC) and its international hierarchy of data archives to the proposed Square Kilometer Array (SKA) radio telescope, which may produce as much as an exabyte of data every few days. This data tsunami is not limited to the physical sciences; the biological and social sciences are being inundated as well.

One of the major lessons from web search and cloud data centers is the power of truly massive scale, near real-time data analysis. When anyone with a cheap cell phone and a web browser can extract data and insights from a non-trivial fraction of the human knowledge base, behavior and culture are transformed. I would like to believe that we can bring the same data-driven analytics to scientific research as are routinely exploited to find a good lasagna recipe. We need a balanced exascale initiative, in every sense of the word, lest we be a technical confirmation of the Sapir-Whorf hypothesis.

That’s One Small Step for {a} Man …

N.B. I have posted additional reflections on the lessons of Apollo for sustaining innovation on the CACM blog.

On Sunday afternoon, July 20, 1969, I rode my AMF Roadmaster bicycle to the local Gulf gas station and asked for their cardboard model of the Apollo Lunar Module. (Some of you will recall that it was originally called the Lunar Excursion Module, whence the common name LEM.) I pedaled home in time to watch the Apollo 11 lunar landing later that afternoon, holding the cardboard model of the LEM in my hands.

It was only years later that I came to understand how perilous that landing had been, due to inadequate lunar images for landing site selection, a critical fuel shortage and an overloaded guidance computer with less power and storage than one of today's engineering calculators. As Apollo 13 later showed, the risks were real. Nonetheless, that day remains burned in my memory and it made no difference to me that the images transmitted from the moon were in black and white, because all we had at home was a black and white television.

I was amazed then, as I am now, that NASA had planned for the astronauts to sleep before venturing to the lunar surface. Fortunately for me, the decided to forgo the sleep period and stepped onto the lunar surface while it was still before my bedtime, as a twelve year old.

It Only Seems Miraculous the First Time

As a budding geek and a voracious reader of science fiction, I was convinced then that the lunar landings were the beginning of humanity's long-term commitment to space exploration. Forty years have proven me wrong. As I noted in Innovation: A Plane of Excellence, my friend Thomas Sterling once posed a question that still haunts me, "In which year of birth (1930 or 1970) would one have had the higher probability of walking on the moon?" We know the answer. What really happened?

We all know the "space race" was a creature of the Cold War. After the Apollo 11 landing, public interest waned quickly, the Vietnam War consumed increasing amounts of the U.S. federal budget, and science had always taken back seat to the race itself. (Harrison Schmitt was the only geologist among the astronauts who walked on the moon.) Many people do not realize that the NASA budget was already in decline at the time of the first landing.

Exploration Touches Something Primal

At the request of the White House Office of Science and Technology Policy (OSTP), a panel chaired by Norm Augustine is currently reviewing NASA's human space flight program, with a six month timeline to produce recommendations. We can all debate the cost/benefit ratios of human and robotic space exploration, and I will be the first to argue for the power and value of automated exploration. However, there's something primal about being there. It touches something deep in our psyches, to explore and see for ourselves. More to the point, we spend more money on far less important things.

In a Sunday New York Times editorial, Tom Wolfe captured this conundrum, noting that NASA lacked – in his words – a philosopher, someone who could capture the passion and thrill of exploration and relate that to the public. It's simple really; it's a hunger older than history, a passion never fully satisfied – the eternal desire to discover and to know. Let's go back, and then go beyond – Mars beckons.