Harry D. Huskey: Obituary

Featured

 

On April 9 of 2017 the Computing World lost one of its pioneers, Harry Douglas Huskey. His work spanned all that an academic, an engineer and a scientist can aspire to: building innovative hardware, novel software, and educating computer scientists to continue his legacy world-wide.

Harry Huskey was born over 101 years ago, on January 19, 1916 in the Smokey Mountain region of North Carolina, but soon moved with his family to Idaho. Being raised on a farm, he was used to dealing with all issues that would arise, a trait his family, colleagues, students and friends valued immensely.  As a high school student, Mathematics was easy for him. He obtained a BS degree from the University of Idaho in 1937.

An application to a Rhodes scholarship led to a Teaching Assistant (TA) position at Ohio University in Athens. There he was involved in a project using electro-mechanical relays for calculations, using three magnetizing coils with 6 contacts to allow storage and readout of one bit. In 2013, when inducted as a fellow of The Computer History Museum, Huskey stated “I didn’t complete it because I decided it was too expensive and there wasn’t any use for it” mirroring a 1939 decision by Bell Labs management. In 1939 Harry married Velma Roeth, a student in his Athens geometry class. Working with Professor Tibor Rado, an émigré from Hungary, he completed a thesis addressing a problem of inscribed polyhedra, stated in 1911 by Z. de Geöcze, A year later Harry obtained a TA-ship at Ohio State University in Columbus, where he received his Masters and then PhD in 1943.

In 1943 Huskey accepted an instructorship in Mathematics at the University of Pennsylvania. He, Velma, and their baby daughter moved to Philadelphia. A second daughter was born and Harry took on outside teaching jobs. When a position at Moore School at the University materialized, in March 1945, he obtained the clearance needed to encounter the ENIAC, a computer intended to compute firing tables for the military.

ENIAC programmed by Harry Huskey.

Those tables used to be computed manually, using 10-digit precision mechanical desk calculators, as seen in the movie Hidden Figures.  On the ENIAC these calculations were performed by its 20 electronic calculators, the controlling programs, initiating calculations and transferring their results were provided through wiring trays, set up for each type of task. Once set up, its speed allowed smaller integration intervals at high speed, which increased the accuracy of the results. Test results published in 1949 by Huskey, pointed out that the 10-digit fixed point capability created serious rounding errors in the earlier methods.

ENIAC Programming Team at work

Storage in the ENIAC was achieved by using about 11,000 of its 18,000 vacuum tubes. Although they were pretested, their reliability caused daily downtimes. Attempting to scale ENIAC would further increase failure rates.  A successor project, EDVAC, was to use mercury delay lines to store data, regenerating data dynamically as the bits circulated, mirroring modern dynamic DRAMs. EDVAC, as proposed by Presper Eckart and John Mauchly was also to have an addressable program storage, now known as a von Neumann architecture. Harry was offered the position to lead the EDVAC project development, and an associated research professorship, however, the chair of the topology-oriented Moore School, withdrew that offer.

Lectures given at the Moore School, using documents written by Harry Huskey, had brought him into contact with many scientists interested in computing, including Derek Lehmer of UC Berkeley, Maurice Wilkes from Cambridge, UK, and Frederic Hartree at the UK National Physics Laboratory (NPL). Hence, Huskey accepted an offer in the fall of 1946 from NPL in Teddington where many scientists that had been associated with the British code-breaking projects were coming together, focusing on Alan Turing’s design for the Automatic Computing Engine (ACE).

Multiple computer development projects had been initiated in the UK. Huskey was able to visit many of them and assess the suitability of their approaches.  A group at the University of Manchester, where Hartree was on the faculty, was working on the `Baby’ — still weighing about 1000 kilograms — as a prototype for their Mark 1. Its major innovation was the use of the Williams tube for storage, initially 128 bits, and eventually 1024 bits per tube. The Williams tube circuit senses secondary emissions when an illuminated spot on a cathode ray tube face is rewritten, allowing regeneration of illuminated spots or letting them fade.

A Williams tube on display at UCSC in 2015.

When it became operational in June 1948, it was the first working von Neumann architecture computer.  Maurice Wilkes’ EDVAC, using that storage technology, becoming operational on March 1949, led to commercial computers.  Progress at NPL on ACE was slow, frustrating Turing. Huskey and colleagues initiated a prototype project, the Pilot Ace. It started running May 1950, more than two years after Huskey’s return to the US.

A brief vacation in California before sailing to England in 1946 had convinced the Huskeys that their future should be in California. When, in the fall of 1947, the US National Bureau of Standards (NBS) announced plans for an Institute for Numerical Analysis (INA) located at UCLA. Huskey applied and was appointed. Huskey returned from England in early 1948, going first to the NBS offices in Washington, DC.  NBS was to monitor the acquisition by the US Census Bureau of computers, input/output equipment, and archival storage to be used for the tabulation of the 1950 census. Progress was slow.  Based on his experience from England, Huskey proposed a computer design for a NBS Eastern Automatic Computer (SEAC).  That design was shelved when it was decided to obtain a copy of the EDVAC.  Getting a computer for the INA in California was difficult, most available computers exceeded the $350K budget and Eckert and Mauchly’s new corporation (EMCC), with their upgraded EDVAC (renamed Univac) was blocked due to concerns that EMCC had hired engineers with communist leanings.

Huskey decided to use the budget and his experience to build a computer at INA, named the Standard Western Automatic Computer (SWAC). It used 2500 vacuum tubes and 4,000 diodes for its logic and 40 Williams tubes, each storing 256 bits, providing parallel access to 256 Words of 37 bits each for data and program storage. Adding two integers and storing them took 64 Microseconds. SWAC became operational in July 1950 and was then the fastest computer in the world, and the first electronic computer in California. In 1953, the NBS branch in California was closed because of congressional pressure, and the INA project was transferred to UCLA. The SWAC remained in use there until 1962. Derek Lehmer from UC Berkeley had used it to compute Mersenne numbers. George Forsyth, later founding the Stanford Computer Department, also came to NBS to use the SWAC and his wife, Alexandra, did the programming.

Harry Huskey at the INA SWAC. The bank of 40 storage tubes (some were spares) with the screens showing, is central.

Around 1952 Huskey had started to refine the design for the SWAC, replacing the storage tubes with a rotating magnetic drum and downsizing the computer so that it could be run by a single person in a regular office, an early concept of the “personal computer”. It was implemented and built by Bendix as the G-15 in 1954.

In 1952, the Institute of Radio Engineers (IRE) asked Huskey to organize a computer group, which was called the PGEC (Professional Group on Electronic Computers). This group became the IEEE Computer Society.  Huskey was editor for PGEC until 1958 and later was editor of the IEEE Transactions on Electronic Computers.  He was vice president (1958) and president (1960) of the ACM during this time.

In absence of a CS department in 1954 at UC Berkeley, Huskey was offered a joint faculty position in Mathematics and Electrical Engineering. Bendix provided a G-15 for development in Huskey’s Berkeley home, so that his family now had a “personal computer.”

Harry Huskey, seated with his G-15, now exhibited the Smithsonian in Washington, DC

Huskey and Joe Weizenbaum, later at MIT developed and added a package for floating point arithmetic. With a reliable computer, Huskey focused on its use and software. Having programming languages became a crucial aspect for computer use. Under a Navy contract Huskey developed NELIAC for its control applications. That experience provided an environment for his students like Niklaus Wirth to develop versions of Algol, PL360, and Pascal. Some other languages like the HH, an interpreter for on line use, and CAL developed by Butler Lampson were research efforts under his guidance during early 60s.

To enable on-line computer access, Huskey obtained ARPA funding in 1963 for Project Genie, incorporating the paging hardware into a SDS930 computer for a multi-user timesharing system. The project was directed by Melvin W. Pirtle and Wayne Lichtenberger under his guidance at UC Berkeley and was later commercialized as SDS940. Butler Lampson and Peter Deutsch were among the students who made significant contributions. David Evans joined him at UC Berkeley on the project after working at Bendix on the G-15. Huskey, when called upon to fix problems in managing the University Computer Center, assigned the task to Evans. Evans soon after founded the CS department of the University of Utah and the Evans-Sutherland computer graphics company.

Harry Huskey went to Kanpur in 1963 to introduce computing to the academics and researchers in India, when USAID decided to initiate an Indian Institute of Technology at Kanpur, to be supported by consortium of five Universities. Kanpur was an unlikely setting; a large city characterized by traditional industries, and only a military airport. An IBM 1620 with its maximum core memory of 40,000 6-bit digits (or 30,000 characters) was procured and transported using traditional Indian transportation.

The IBM 1620 arrives at IITK

Harry checks the IBM1620.

He recruited other faculty from leading US Universities, as Foreman Acton (numerical analysis) and Irv Rabinowitz (systems) from Princeton to join him to train aspiring computer scientists in India and neighboring countries. Faculty from other IIT Kanpur departments and projects were engaged as well. Soon 10-day courses were offered, mirroring the means and effects of the 1946 Moore School courses. Over 2000 external scientists were trained over the ensuing 7 years. His academic successor was H.K. Kesavan and Gio Wiederhold covered systems. Late in 1964 Harry Huskey convened a Ford Foundation sponsored conference and advisory meeting at Kanpur, bringing ten leading computer scientists for a week to Kanpur, putting IIT Kanpur on international CS map. Huskey also initiated the Indian Computer Users Group that morphed into the Computer Society of India (CSI) in 1964. Many Indian computer scientists and the IT business in India trace their lineage to these efforts.

In July 1964 Huskey returned to UC Berkeley. In 1967 he joined UC Santa Cruz (UCSC), as the founding father in developing a computer science curriculum, even though there was no engineering school at UCSC until 1971. He recruited promising young Computer Science faculty to staff the UCSC CS initiative. Continuing after his retirement from University of California in 1986, he accepted invitations from many institutions for decades, allowing him to spread his insights and inspiring computer professionals worldwide. While Harry Huskey received many awards throughout his life, his peers and students recognize his contributions to computing to be broader than those awards could recognize.

Velma Huskey died 1991, ending their joint project – a biography of Ada Lovelace. Three years later he married Nancy Whitney-Grindstaff. They settled in South Carolina, travelled widely, returning to Santa Cruz in 2011. Nancy died in 2015.  Harry left three daughters and one son, and grandchildren, all in the Santa Cruz, CA area.

***********

Gio Wiederhold, Bijoy Chatterjee, and H. Douglas Huskey, May 2017.

Sources

William Aspray: Oral History of Harry Huskey; Spartansburg, SC, for the Computer History Museum, CHM Reference number: X3455. February 7, 2006.

Gordon Bell, et al.: Computer Pioneers: Pioneer Computers Part 1; Computer History Museum, 1996, YouTube v=qundvme1Tik,

Gordon Bell: Mauchli, Maurice Wilkes, Computer Pioneers: Pioneer Computers Part 2; Computer History Museum, YouTube v= wsirYCAocZk,

Harry D. Huskey: Harry D. Huskey, His story; Book Surge, 2004.

Pamela McCorduck: Oral history interview with Alexandra Forsythe; Univ. of Minnesota, Media Archive, May 16, 1976 [SWAC & NBS at 24-31 min].

Vijay Rajaraman: History of Computing in India 1955-2010; IEEE Computer Society History Committee, 2012.

Sam Roberts: Harry Huskey, Pioneering Computer Scientist, Is Dead at 1001; New York Times, 20 April 2017.

E.C. Subbarao: An Eye for Excellence; Harper & Collins, 2008.

Gio Wiederhold: report and photographs of the USAID KIAP project, 1964-1965, http://i.stanford.edu/pub/gio/personal/1965India

Wikipedia articles on Huskey, SWAC, Pilot ACE, IIT Kanpur, ACM past presidents, …

F.C. Williams et al.: Manchester Baby and the birth of Computer Memory; YouTube v=SpqayTcGcw, 2013.

Remarks at Harry Huskey’s funeral service; 15 April 2017.

 

—————————- o ———————————– o ————————–

Computers: The Early Years

Dr. Harry Huskey was was born in the Smoky Mountains region of North Carolina and grew up in Idaho. He gained his Master’s and then his PhD in 1943 from the Ohio State University with a thesis on Contributions to the Problem of Geocze.

Harry Huskey at the SWAC

Harry Huskey taught mathematics at the University of Pennsylvania and then worked part-time on the early ENIAC computer in 1945.

Harry Huskey designed and managed the construction of the Standards Western Automatic Computer (SWAC) at the National Bureau of Standards in Los Angeles (1949–1953). He also designed the G15 computer for Bendix Aviation Corporation, which could perhaps be considered as the first “personal” computer in the world. He had one at his home that is now in the Smithsonian Institution in Washington, D.C.

Computer History Museum Honors Dr. Huskey

Computer History Museum 2013 Fellow

Myanmar’s First Computer