Chen-Bin Su

From SEG Wiki
Jump to navigation Jump to search
Chen-Bin Su
ChenBinSU headshot.png
Latest company Parallel Data Systems

Chen-Bin Su and Wes Bauske founded Parallel Data Systems, Inc. in 1997. PDS provided the oil and gas exploration industry with high-quality 3D prestack depth migration with aff ordable cost and quick turnaround time. Su and Bauske, started the company with a small amount of capital, taking a high level of personal risk. Zhiming Li subsequently joined them, leaving behind a secure job with a major oil company. Th e business was a bootstrap operation for the fi rst three years. It turned into a US$17 million business in 2006. TGS, which bought PDS, was impressed by its ability to solve highly complex imaging problems on massive data sets while meeting very challenging project schedules. Th ey continually increased throughput capacity and expanded their imaging algorithms, as well as commercializing the next generation of seismic imaging technologies including wide-azimuth 3D.

Biography Citation for the Cecil Green Enterprise Award

Contributed by Rutt Bridges

Parallel Data Systems, Inc. (PDS) was founded by Chen-Bin Su and Wes Bauske in 1997. Starting with a small amount of capital, they worked initially in their garages, then their living rooms, and then fi nally moved into a small rented offi ce space in 2000 after completing their fi rst contracts. Su and Wes were classic entrepreneurs, taking great personal risks to achieve a dream.

They had both been involved in R&D and highperformance computing since joining Geophysical Services Inc. in 1979. Later they worked together for IBM doing R&D in seismic parallel computing environments from 1988 until 1997. In 1992 Chen-Bin Su was honored for the “Best OTC Geophysical Paper” for his work in networked cluster parallel seismic computing. Cluster parallel computing was a new concept in those days, when the industry’s attention was on supercomputers or massively parallel machines. In 1997, they decided to make this pioneering idea a commercial reality by founding PDS in Dallas. In those days, prestack depth migration (PSDM) was the “grand challenge” of the seismic industry. Their vision was ambitious: large-scale PSDM on clusters of reduced instruction set (RISC) workstations.

Two years may seem like a relatively short time but, as any entrepreneur will attest, it is a long time to be writing software with no income. As Einstein said, “Put your hand on a hot stove for a minute, and it seems like an hour. Sit with a pretty girl for an hour, and it seems like a minute. That’s relativity.” Su and Wes were on a hot stove. The work is intense and families have to be patient and believe in the dream. Finally, in 1999, they had a working cluster-parallel PSDM and they landed their first job, a PSDM project for Unocal in the Gulf of Mexico covering nine OCS blocks (90 square miles/230 square km).

As was always the case for Wes and Su’s code, their cluster PSDM was fast. However, it was arguably not as elegant and precise as some slower versions offered by competitors. That changed after Zhiming Li joined the company as the third owner in 2002. Zhiming was one of the industry-recognized thought-leaders in PSDM. Zhiming’s scientific talents combined with the computational skills of Wesplus Su’s ability to always find a way to succeed combined to significantly impact on the success of the company.

In 2002 PDS also grew by opening an offi ce in Houston. They continued to address the “grand challenge” of how to perform larger and larger 3D PSDMs quickly, economically, and accurately. PDS expanded dramatically both in staff and computer power from 2002 through 2007. The company had 12 IBM RS/6000 workstations in 1999; by 2007 it had a 6,000+ CPU cluster for more iterations and higher quality results. PDS was sold to TGS-NOPEC in 2007 for US$72.5 million. At that time PDS had 21 employees and $17.5 million in revenue with $11.7 million in profits. The company had always stuck to its vision: “Provide the oil and gas exploration industry large volume and high quality 3D prestack depth migration with affordable cost and quick turnaround time.”

Remarkably enough, PDS never had a dedicated marketing staff. They believed that, if they delivered high quality PSDM on time and within budget, sales would follow. Satisfi ed customers were their best sales and marketing staff, and better still they were free. The company maintained a fairly consistent one-year backlog. Given what Cecil Green Enterprise Award stands for, Parallel Data Systems, Chin-Bin Su, Zhiming Li, and Wes Bauske are outstanding representatives of the qualities embodied in this award.