Ask HN: Is client computation a viable alternative to ads?
Would it be a reasonable business model to have visiting users donate computational power (via the javascript vm in their browser) as long as they're browsing a site?
An possibly valuable and easily parallelizable task is brute forcing (mid to low level) encryption keys.
How many page impressions add up to the equivalent of an hour of EC2 time? How does that compare to ads?
How does it make money?
Distributed computational resources have a value. For example, there is a list of scientific projects being conducted via many client computation: http://en.wikipedia.org/wiki/List_of_distributed_computing_p...
No of course, but those are research projects. I assume you're talking about a business, which implies commercial. Is the govt going to fund your business like they (directly or indirectly) fund these research projects, or will there be paying clients/customers? Who will need to crack encryption keys (or whatever..) often enough to pay you instead of setting up their own cluster? Ads make money because they make the advertisers money. Is searching for gravitational waves, or ET going to make someone money?
Commercial opportunities include protein structure prediction for drug design, brute-force attacks for breaking security, and various simulations (geophysics, finance).