NSDI is the abbreviation for the USENIX Symposium on Networked Systems Design and Implementation. It’s a highly regarded conference for Systems researchers. I’ve been occasionally scanning the proceedings for 2017, reading a paper here or there.
Some of the folks at the Stanford DAWN project attended the 2017 meeting and wrote up their perspective. Definitely provides a different angle from the way I was looking at the conference proceedings:
A group of us at DAWN went to NSDI last month. The program was quite diverse, spanning a wide variety of sub-areas in the networking and distributed systems space.
We were excited to see some trends in the research presented that meshed well with the DAWN vision.
In bullet points the trends were:
- More support for machine learning
- Video as a data source for analytics
- Embracing the use of hardware accelerators and FPGAs
- Frameworks that exploit fine grained parallelism
- High performance with high programmer productivity
DAWN is shaping up to be an interesting project, in the Berkeley CS tradition of highly collaborative research teams bounded by a a 5 year lifespan. Go figure, given some of the principals involved.
Also, Matei Zaharia and Peter Bailis popped up on an ArchiTECHt podcast, which was pretty informative. Alex Ratner had a related discussion with Ben Lorica on generating training data with limited resources.