I'd like to hear about hardware startups and your troubles and problem solving. Particularly if there are any IC manufacturers or fabless startups.
This is an area (EE and ICs especially) I was always fascinated by, but I have no practical knowledge about. Yet, I have several ideas which I would like to develop into a HW product (one of million ideas we all write down), but I have no one to talk about since I don't know any people that are experts in. My exposure to the area comes mostly through occasional scouring through IC manufacturers news/prs/blogs, eetimes and wiki (so it's nothing basically).
Is it possible at all for a small yc-like startup to break into an IC field at all (fabless like Nvidia or manufacturer area). Costs associated seem to be astronomical.
Octopart is a YC-funded search engine for electronic parts.
One reason why I'm listing these indirectly-related examples is that hardware for me seems to involve an ecosystem of different companies, each painting part of the overall picture. To me it's a very inter-connected industry: from the hardware manufacturers to the design houses to the software tool providers to the customers, etc.
As you said, manufacturing and selling a hardware product requires a sizable capital investment, not to mention challenges of maintaining a logistical operation. If the product design is not too complex, one can get it manufactured for cheap in Taiwan or in China... startups based somewhere else will just have to find the nearest board house to do the first prototypes.
Fabless is of course a good way to go as it eliminates a lot of the costs. That also brings with it business-model-related questions: do you sell the IP or do you actually want to make something in the end? Are you thinking about opensourcing your IP? Etc...
There are hardware-as-a-service startups, opensource hardware startups like Arduino(http://www.arduino.cc/), and many such companies coming into the market in recent years. It's definitely a rich playground with decent barriers to entry(hardware expertise, market knowledge, connections) and I feel that there isn't a "killer" business model out there yet for smaller players to capitalize on.
Would be interested to know what other hardware-related startups are doing!