Download presentation
Presentation is loading. Please wait.
Published byClinton Watkins Modified over 9 years ago
1
Research Issues Sachin S. Sapatnekar ECE Department University of Minnesota 1
2
Setting a research agenda Questions to ask yourself –What are my key skills? Basic skills: Algorithms, design knowledge, probability theory, … Areas: high-level synthesis, physical design, DFM, … –What’s important/interesting? What would I like to do? What’s hot? And likely to remain so? –Collaborative opportunities –How easy is it to “jump” to an allied area? Pros vs. cons Don’t jump too often! Balance all of these appropriately 2
3
The life cycle of a new idea Inception –A few “early adopters” pushing the idea Adolescence –A number of people begin to address the problem –Proposals on the topic get funded –Sessions on the topic at conferences; “early adopters” in demand for invited talks Youth –Tons of papers on the topic –But funding isn’t “automatic,” low-hanging fruit have been plucked Middle age –Tons of papers on the topic (top conferences begin to get bored with it) –Funding becomes much harder –Can still make a strong contribution, though it’s harder Retirement –Remember channel routing? Reincarnation –The same idea becomes applicable in a new field, in a different way –The twentysomethings think they’ve invented new algorithms; the sixtysomethings sigh 3
4
Publishing paradigms Two ways of making an impact –Define a new problem –Devise a clever solution to an old problem Tread carefully –Working on what’s hot at DAC/ICCAD/… Timing is key –Working on a brand new topic that nobody has tried before Risk vs. reward 4
5
An incomplete list of concrete ideas READ. READ. READ. –Corollary: Volunteer to review papers Keep publishing –Funding is something you can’t control; publishing (to some extent) is Talk to your colleagues, even in areas distant from yours –Sit in their classes, have lunch with them, … Tried something and it didn’t work out? Learn when to move on –Let (some) ideas slow-cook in your head Consciously learn to multitask, if it doesn’t come naturally FOCUS. FOCUS. FOCUS. 5
6
Recruiting students Undergraduates –You’ll find several who want research experience –Chances are, you may not get a (CAD) paper with them There are exceptions to this –But you’re planting a seed Check for systemwide opportunities in your university –WISE, REU, … Grad student recruiting –Sight-unseen vs. try-them-out-first –Acing exams != doing research – GPA alone is not enough 6
7
Ties with industry See how your work can be applied in industry Making contact –Don’t be shy: talk to people at conferences, invite yourself over to their companies The miss ratio is high The time constant/delay can also be high –Industry is a good source for problem statements, internships, benchmarks, funding, … Building relationships takes time: be patient 7
8
Students - friends or fodder? Your students are your greatest asset –Nurture them –Support them –… but don’t spoil them! e.g., make them think: resist the urge to solve problems for them Set clear guidelines and expectations on standard issues –Communication: Meetings/email/reporting –Writing standards Be flexible wherever possible –Recognize that each student has his/her own style of working, which may not be the same as yours –Recognize their aptitudes, and tailor their tasks accordingly Learn from them Leverage your classes where possible 8
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.