As part of the SEED project, I recently talked with Timothy Tsai, a Senior Research Scientist at Nvidia. In this interview, Timothy shared with me that an engineer in industry must be able to develop new ideas and discern the bigger picture of a technical project. Timothy also explained that technology companies must pivot to best engage with new markets and opportunities. Read Timothy's responses to the three SEED questions to learn more about his unique perspective on the engineering of software.
As with many successful companies, Nvidia has undergone several transformations in its focus. Originally it produced chips for graphics processing. It then formalized the functionality and interface to the graphics hardware and software to invent the GPU. After several years, GPUs started being used for efficient parallel processing for supercomputers. In the last few years, the world realized that GPUs offered a practical mix of parallel processing units and memory bandwidth that enabled efficient deep neural net (DNN) implementations. So, in 25 years, Nvidia has transformed from a graphics company to also being a company that powers self-driving cars.
Finding out how things are implemented. This includes (1) finding the right experts who also happen to have spare time and (2) having access to a search engine (i.e., both the Internet and your company's or school's intranet) that actually finds pertinent resources.
Try to understand the landscape of the projects that you are working on beyond just the expertise needed for those specific projects. Try to think about new ideas and extensions. This striving for new ideas will make those projects better, increase your understanding and experience, garner notice from your bosses, and make your career an interesting endeavor instead of merely a way to make a living.
Do you have any questions about this interview? Or, do you have a response to Timothy's ideas? If you do, then I hope that you will contact me to share some of your thoughts. Or, do you want to be updated when I publish new interviews in this series? If you do, then please subscribe to my mailing list.
Enjoy this post? If so, please read, Using real faults to evaluate test suite prioritization techniques, my most recent article.