Published by
Stanford Medicine

History, Humor, Medicine and Literature, Science, Technology

Half-century climb in computer's competence colloquially captured by Nobelist Michael Levitt

Half-century climb in computer's competence colloquially captured by Nobelist Michael Levitt

ancient computerOn October 9, the day Stanford macromolecule-modeling maven Michael Levitt, PhD, won his Nobel Prize, I wrote him a note of congratulations.

He wrote back six days later: “Thanks so much. It has been one wild ride! It will be good for the field, though, and I will learn to disappear and still have time for myself.” It’s a wonder he got back to me as soon as he did, crushed as he must feel by the cheering throngs dogging him at every turn since Nobel day. But he has made a point of replying quickly and gracefully to not only well-wishers but deadline-driven reporters.

Although his Nobel was for chemistry, the lab Levitt operates in is stocked with shelves full of ones and zeroes. His expertise lies in the field of computer science, a field of which Alfred Nobel had no inkling when he created the awards in his final will, written in 1895.

As we all know, Nobel made his millions in the explosives field. No explosion he could have imagined in 1895 has been more profound, in recent decades, than the explosion in computing power pithily encapsulated in Moore’s law. In the late 1960s, Levitt began constructing his increasingly detailed simulations of the giant biomolecules that animate our cells and, in a sense, our souls as well, by pumping punchcards into what was then among the world’s most potent computers (dubbed Golem in memory of a powerful, soulless giant of medieval Jewish folklore) at Israel’s Weizmann Institute.

Since those seminal days, the ones-and-zeroes game has picked up speed. Responding to an e-mailed query from science writer Lisa Krieger of the San Jose Mercury News, Levitt put it this way:

The computer that I used in 1968 allowed me 300 [kilobytes] of memory, or about 1/10,000th of the memory on a smart phone. [An extremely complex,  fifty-step computation] took 18 minutes on the Golem computer… for a cost of about five million 1965 U.S. dollars ($35 million today). The same calculation takes 0.18 seconds on an Apple MacMook PRO laptop costing $3,500. This means that the calculation is… 6,000 times faster on a computer costing… 10,000 times less.

If cars had changed in the same way, Levitt drolly noted, “a 1965 Cadillac that cost $6,000 in 1965 dollars ($40,000 today) would actually cost just four dollars. More amazingly, it would have a top speed of 600,000 miles an hour and be able to carry 50,000 people.”

Makes me wonder: Just how long will it be before we can no longer tell our computers from ourselves?

Previously: But is it news? How the Nobel Prize transformed noteworthy into newsworthy, Nobel winner Michael Levitt’s work animates biological processes, No average morning for Nobel winner Michael Levitt and Stanford’s Michael Levitt wins 2013 Nobel Prize in Chemistry
Photo by kalleboo

Comment


Please read our comments policy before posting

Stanford Medicine Resources: