Nvidia May Change Direction of GPU Development as It Switches Chief Scientist

Nvidia Replaces Veteran David Kirk with Stanford Scientist William J. Dally

by Anton Shilov
01/28/2009 | 03:29 PM

Nvidia on Wednesday said that William J. Dally, the chairman of Stanford University’s computer science department, will join the company as chief scientist and vice president of Nvidia research and will replace the veteran chief scientist of Nvidia David Kirk. The switch of the persons who are not responsible for rapid and general business decisions may indicate that the company may change direction of its chip development.

Long Time Graphics Guru Leaves His Post


David Kirk, who is leaving his current position to become an Nvidia fellow and who has been Nvidia’s chief scientist for about twelve years, has a long history as a graphics technology developer and visionary. Mr. Kirk is the inventor of 50 patents and patent applications relating to graphics design and has published more than 50 articles on graphics technology. In 2002, Dr. Kirk received the SIGGRAPH Computer Graphics Achievement Award for his role in bringing high-performance computer graphics systems to the mass market. Prior to Nvidia Mr. Kirk worked at Apollo Systems Division at HP on commercial RISC microprocessors and later on worked at a video game company.

“I would also like to congratulate David Kirk for the enormous impact he has had at Nvidia. David has worn many hats over the years – from product architecture to chief evangelist. His technical and strategic insight has helped us enable an entire new world of visual computing. We will all continue to benefit from his valuable contributions,” said Jen-Hsun Huang, president and chief executive officer at Nvidia.

Parallel Computing Expert with Ongoing Microprocessor Project Becomes Chief Scientist of Nvidia

Bill Dally, unlike his predecessor, is in expert in parallel computing technologies as well as interconnection networks, not graphics. According to his biography at Stanford University’s web-site, Mr. Dally worked with Cray since 1989 on the development of many of their supercomputers including the T3D, T3E and Black Widow; besides, he is currently collaborating with Cray on their HPCS effort. Additionally, Dr. Dally is a cofounder of Velio Communications and Stream Processors companies. Mr. Dally has published over 200 papers, holds over 50 issued patents, and is an author of the textbooks, Digital Systems Engineering and Principles and Practices of Interconnection Networks.

At Stanford University, Dally has been a Professor of Computer Science since 1997 and Chairman of the Computer Science Department since 2005. Most recently Mr. Dally was one of the key persons in Pervasive Parallelism Lab (PPL), an effort of several tech industry giants (Advanced Micro Devices, HP, IBM, Intel Corp., Nvidia Corp., Sun Microsystems) with Stanford University aimed to develop dedicated parallel computing platform.

It is clear that William J. Dally is an expert in parallel computing and has little background in commercial companies. Moreover, based on open facts, Bill Dally has never worked in graphics processing or video game industries.

Obviously, there is a convergence between various parallel processors, such as graphics processing units, central processing units and other; hence, it is crucial to create parallel computing platforms that can be easily programmable. But the most intriguing part is that one of the current projects of Bill Dally at Stanford is creation of ELM (efficient low-power microprocessor), which resembles Intel’s Larrabee or Cell processors in at least some ways.

The ELM project focuses on the creation of a low-power, high-performance, programmable embedded system. The ELM system is comprised of many simple tiles, called Ensembles, that contain software managed memory (EM) and several Ensemble Processors (EPs). By designing a chip comprised of many efficient tiles (processing cores with supporting hardware) and providing a full software stack, ELM will be able to reduce or eliminate the need of fixed function logic blocks in passively cooled systems.

“The Passion for the GPU’s Evolution into a General Purpose Parallel Processor”

In fact, Nvidia makes no secret that Mr. Dally was hired because of his expertise in parallel computing and not because of graphics technology development background. What is interesting is that Nvidia openly admits that it is working to transform graphics processor into general purpose processor.

“I am thrilled to welcome Bill to Nvidia at such a pivotal time for our company. His pioneering work in stream processors at Stanford greatly influenced the work we are doing at Nvidia today. As one of the world’s founding visionaries in parallel computing, he shares our passion for the GPU’s evolution into a general purpose parallel processor and how it is increasingly becoming the soul of the new PC. His reputation as an innovator in our industry is unrivaled. It is truly an honor to have a legend like Bill in our company” said Jen-Hsun Huang.

It is noteworthy that Raja Koduri, chief technology officer of ATI, graphics products group of Advanced Micro Devices, is also known for his expertise in the field of GPGPU.