[GTO]

Originally Posted by HPP

True, but location of servers, and server technology has a play in speed of results etc as well(Although those are secondary to the actual quality of results). So in a sense you are right, but the ability to have mass amounts of servers strategically placed throughout the world definitely helps the user experience.

I don't think server location matters wrt to speed & efficiency. With the internet having become what it is today, location is irrelevant. Team-BHP's box is located in California, but I haven't heard anyone complaining about load times (except when there's too many users causing server stress). We also have a backup server located somewhere else on the planet. I can't disclose the location

Originally Posted by srijit

They have to be scalable at the beginning itself, otherwise it wont work.

Scalable yes. But my point was : Their current traffic can be expected to be far (and far) less than 1% of what Google enjoys. Thus, they don't need the infrastructure that google has. Server strength should prolly be capable of serving 25% more users than you expect at any given time.

[srijit]

Originally Posted by GTO

I don't think server location matters wrt to speed & efficiency. With the internet having become what it is today, location is irrelevant. Team-BHP's box is located in California, but I haven't heard anyone complaining about load times (except when there's too many users causing server stress). We also have a backup server located somewhere else on the planet. I can't disclose the location

GTO, that is not a fair comparison. TBHP probably has only one database (which is not replicated). Google has a HUGE distributed database. They also need to have multiple servers so that they can do load balancing and also allow for local searches. In that case, it makes sense for search engines to have multiple servers and distributed data so that there is no single point of failure. Am sure both MSN and Yahoo also do the same.

Originally Posted by GTO

Scalable yes. But my point was : Their current traffic can be expected to be far (and far) less than 1% of what Google enjoys. Thus, they don't need the infrastructure that google has. Server strength should prolly be capable of serving 25% more users than you expect at any given time.

That would have been OK, if they hadnt done the advertising that they did. They billed themselves as a better product than Google, said that they were run by ex Top Google Engineers and just created a lot of hype. I'd say that they got atleast 10% or more of Google's traffic.
But the problem was their interface was not that good and their search results were rubbish. That will immediately cause people to go back to Google.

The simple fact is, if you want to beat Google (or any other company), offer something that they dont have and do it better. Not just equally good, but better.