Enter the secured room in the Schorr Center that houses Sandhills and Red, the two supercomputers on the University of Nebraska-Lincoln campus, and you meet with cold air and a low hum.
The buzz of the machines muffles some of what David Swanson, director of the Holland Computing Center, says as he walks the length of the Red setup. But the scores and scores of stacked two-terabyte hard drives offer visual evidence of a discussion involving words like teraFLOPs, petabytes, Large Hadron Collider data storage -- these supercomputers can cook.
The cores that make up these supercomputers seem to improve exponentially. In fact, Swanson said, the Holland Computing Center's annual budget is based in part on that.
"To date, and for the next several years, it appears that something called 'Moore's Law' will hold," he said. "Engineers will succeed at doubling the number of transistors that can be laid down in a given area every 18 months. The consequence is that computing will continue to improve exponentially."
The capabilities of computers may have catapulted from one era to the next, but the climate and the hum in the supercomputer's home mirror a similar area on the second floor of Nebraska Hall about 55 years ago for a first-of-its-kind space on the UNL campus, a computer center.
The floor was raised several inches to hold reams of wires strung along the base. The air conditioning was set beer-cooler low, to combat the heat that would pour off vacuum tubes. The same collection of consoles, key punch card cabinets and cables inside Nebraska Hall appeared on the "Lost in Space" spaceship and inside the Adam West-era Batcave.
It was home to UNL's first digital electronic computer, a Burroughs 205, also known as a Datatron. And even though it had all the processing power of a modern-day hand calculator, it foreshadowed the promise of what computing could become, said Don Costello, who was one of the first to use UNL's.
"I arrived in Nebraska on the first week of September (1960) and the machine was put in on the second week,” said Costello, a former Air Force intelligence officer and IBM employee who enrolled at UNL as a doctoral student in mathematics. "I came and I said, ‘Do you have any computers on the campus, and this guy said, ‘Yeah, we got a new computer!' I went over right away and started working on it.”
Costello, now an emeritus associate professor in UNL's School of Computer Science and Engineering, is working on a book about the early days of computing at the university. He plans to write of the men and women who propelled the research institution toward its supercomputer status. It starts chronologically, with the B205 era.
The first digital computer cost $180,000, but thanks to a grant from the National Science Foundation and a 50 percent education discount, UNL didn't spend a dime on the parts. The labor, however, might have been a different story.
A 100-by-30-foot space on the second floor of Nebraska Hall, the building that then also housed the Elgin Watch Company, was marked off for its installation. The Datatron's main cabinet weighed 1.5 tons. Another seven cabinets came along with that one. In the Daily Nebraskan's Feb. 2, 1960, article about the computer purchase, this machine was described as "medium sized."
"This new Center (sic) will open areas for research activity throughout the University," Dean John C. Weaver told the Daily Nebraskan in February of 1960, when the purchase of the computer was first announced. "In many research areas, a man with a pencil and paper can't make much headway."
Costello said the few who used it at first had to learn how to load paper into it, change tubes, input data by way of stacks of punch cards. A "0 4259 64 4955" can represent "BRUIN," according to B205 instruction manual, which you can find online, along with an emulator that allows present-day B205 fans -- they exist -- to run programs on a digital version of the archaic device.
"There was quite a bit of preparation and a lot of futzing around till we got things going," Costello said. "But as soon as we got that Burroughs 205 going, it was a very reliable machine after that."
A day with the Datatron was a formulaic day, Costello said. The nature of computing in the early '60s didn't allow for much multitasking.
“I would say, in my mind, it was manageable," he said. "I wrote a flow chart. I wrote out the code. Then I went to the key punch machine. Then I put the code in. Then I made mistakes in both syntax and logic. Then I had to fix that. Then I got my data ready. Then I ran my data. And there was a break in-between everything. I did something, then I'd have coffee or call my wife or I'd write my term paper or whatever the hell I was doing and then there was a break. To me, it was more comfortable, but it took me forever to get something done."
Dr. Gordon A. Gallup, then an assistant professor of chemistry, told the student newspaper in 1960 that the computer would be used to solve only two kinds of problems -- statistical analysis and scientific calculations. But programs created elsewhere on the Datatron showed that plenty could be explored within those two fields.
In 1956, two mathematicians inputted a set of six musical rules they created based on Top Ten radio and Mozart compositions into the Datatron, got a Tin Pan Alley-style set of notes outputted back to them and announced to the world that a computer could crank out 1,000 pop songs an hour. CalTech programmers taught another machine how to play by-the-book blackjack in 1957.
Another "giant electronic brain" was fed three previous seasons' worth of ballgame and ballplayer data and, upon crunching it, declared that Detroit and Brooklyn would meet for the 1956 pennant.
"What're we supposed to do?" New York Yankees catcher Yogi Berra told This Week after the machine said his team would finish third. "Choke up over a glorified pinball machine? We got the pitchin', we got the hittin'. The only way Detroit could win the pennant is if the machine could pitch for it."
(The Datatron was half right. Brooklyn made it to the World Series. But so did Berra's Yankees.)
In Lincoln, the announcement of the computer purchase included a host of potential uses for the school's first digital computer, many of them related to agricultural studies. Researchers could use it to study the effects of weather on crops. (There were about 400,000 punch cards with Nebraska weather data ready to be analyzed once the machine was installed, according to the DN.) They could analyze the income, expenses and taxes of thousands of farm operators. Soil fertility, swine breeding, poultry genetics -- those could be studied anew as well. They planned to look at all of that and analyze cosmic rays, too.
For Costello, one project that stuck with him from the early days was headed by Don J. Nelson, a retired UNL computer science and engineering professor who was among the first class of inductees in the Nebraska Hall of Computing.
“In 1963, we did what's called a law information retrieval system,” Costello said. “Harvey Perlman was very involved in that when he was a graduate law student. And a guy name John Gradwohl, who was a law professor, and Don Nelson (who also served as the first director of the UNL Computing Center) headed that."
Under Nelson's leadership, the team embarked on a humongous programming task -- inputting all nine volumes of Nebraska state laws into machine-readable form, something that could be searched on the computer by keyword "for almost any information, in a matter of minutes," the Sunday Journal and Star reported.
“We've got nine volumes of the statutes, and we tear them in half," Costello said. "And we hire all law students' wives, and they key punch every sentence in the book. This is called the corpus -- the corpus was the statute law of Nebraska and it had to be right. So then they re-key punched every card and it put a notch if it was right, if the entry was correct. So that made sure that we had every column, every comma, every semicolon. This at that time was a big deal.”
They went on to create a bill-drafting system, which allowed legislative staff to find and replace words or alter statutes as amendments were made.
What Costello, now a lecturer on intellectual property management, didn't realize then was that in the process they might have created a precursor to Microsoft Word.
"I was a computer guy," Costello said. "When I thought of the word 'computing,' I thought calculation. I didn't think of text management. It wasn't a part of my background up to that time. All of a sudden now, it becomes more."
"We invented an online word processing system at the time, Don Nelson did, and it was called Legiform. If we would have been smarter about patents and copyrights and all that, we could have done something with that. But we weren't so smart. We weren't thinking. We didn't think about all that stuff much. We sort of knew but not very well.
"We started to do things. We had a woman who was working on her thesis from London. Back then, working on her thesis and using Legiform to write her thesis and send it to Nebraska.
"That was big stuff in those days," Costello said. "It was the beginning."