Taiwanese Research : My Thoughts
Hello Readers,
I’m Sethupathi Balakrishnan. I’ve worked for Qualcomm. I’ll be sharing some of my thoughts about the research culture and the rigour in Taiwanese Universities. I’m from Semiconductors domain (specifically Physical Design & EDA). Hence, my ideas/thoughts would be around it.
The Taiwanese are probably the best in Asia in semiconductors research. They are just a tiny bit behind the US. Why do I think so?
Here are some of the reasons.
- The academic/research culture is very strong and so are the incentives to innovation/research. NTU-Taiwan has produced so many engineers and given to TSMC/UMC/Synopsys etc. Their academic standards are insanely high and would beat even US standards at times. For eg. I did my post grad from VIT in VLSI Design. I just had one small course on “CAD for VLSI” where they teach the working of various algorithms that are being used in placement/routing etc. But, when I saw the course catalog of NTU, I was shocked and intimidated TBH. The MS students are asked to design Placement/Routing Engines and not only that, they are also asked to improve the existing methodologies, participate in very high quality contests like ISPD Routing Contest , ICCAD Routing Contest, Tau Timing Contest and publish papers. VLSI-EDA is probably the hardest of all the domains in CS/VLSI. It took me about 3–4 years of experience in the VLSI industry to get to this level of understanding.
- Close to the Industry — All activities in the academia are strongly correlated to Industry problems. In-fact it’s so close that gap between the academia and the industry is almost NILL.
- A lot of smart people pursuing their MS/PhD’s in the US and coming back and settling down in Taiwan thereby transferring huge value.
Dr. Yao-Wen Chang is a classic example. I have a strong feeling that this gentleman is single handedly responsible for increasing Taiwan’s GDP and economic strength.
Yao-Wen has contributed immensely not only to Taiwan’s development but to the entire EDA community. He is the president of IEEE-CEDA. He has trained numerous of students who have later become Managers/Section Head’s at TSMC/UMC/Synopsys/Cadence etc and academic heads in Taiwanese institutions. I’ll cite a few of them
- Tsung-Yi Ho (Heads the THETA Lab at NTHU)
2. Iris Hui-Rui Jiang (Heads Iris Lab at NTU)
3. Huang-Yu Chen (Manager at TSMC)
4. H.C Ou (Manager at TSMC — Heads the Digital Design Flow Team)
and many more.
— — — — — — — — — — — — — — — — — — — — — — — — —
4. Strong Mathematical/Physics/Biology Foundations and the research institutions promoting/nurturing research in basic sciences and technology. For eg.
- Academia Sinica
2. ITRI
— — — — — — — — — — — — — — — — — — — — — — — — —
5. Finally — Smaller Country. Taiwan is very small, in-fact it would be smaller than some states in India. It’s a known matter of fact that smaller the country, easier it is to excel. eg. Taiwan, Hong Kong, South Korea, Singapore etc.
Con — The barrier to entry is relatively lower when compared to the US.
I am sure there are many other reasons why Taiwan is so good at Semiconductors. I have just listed some of my key observations.
— Sethupathi Balakrishnan