Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Turning web traffic into a Super Computer

ben.akrin.com

Post reply on HN
Print view View original

11–20 of 45 posts

Re: Turning web traffic into a Super Computer

#11

Post by sosborn »

NegativeLatency wrote: post #10
matte_black wrote: post #6

Earlier quoted context omitted.

AWS at least already has very cheap ways to get compute time if you don’t particularly care about having uniform availability of it in the form of spot instances. The economics of people letting you use their browser doesn’t make sense. The pay is crap and the time when they are browsing is exactly when they would not want their computer doing a bunch of background processing. You’re better off letting people rent ou…

Seti@home has been doing this for years

Almost 20(!)

Re: Turning web traffic into a Super Computer

#12

Post by noahdesu »

This is a great direction for a certain class of problems, but to call this a general approach for building a super computer ignores the biggest challenge of all: communication performance.

The leadership-class super computers have large budgets for network interconnects like Infiniband, because the hardest problems require ultra low-latency, high-throughput RDMA message passing between processes.

Re: Turning web traffic into a Super Computer

#13

Post by matte_black »

NegativeLatency wrote: post #10
matte_black wrote: post #6

Earlier quoted context omitted.

AWS at least already has very cheap ways to get compute time if you don’t particularly care about having uniform availability of it in the form of spot instances. The economics of people letting you use their browser doesn’t make sense. The pay is crap and the time when they are browsing is exactly when they would not want their computer doing a bunch of background processing. You’re better off letting people rent ou…

Seti@home has been doing this for years

Yup

Re: Turning web traffic into a Super Computer

#15

Post by ryanlol »

matte_black wrote: post #6
bob_theslob646 wrote: post #4

Earlier quoted context omitted.

Why should we keep dreaming?

AWS at least already has very cheap ways to get compute time if you don’t particularly care about having uniform availability of it in the form of spot instances. The economics of people letting you use their browser doesn’t make sense. The pay is crap and the time when they are browsing is exactly when they would not want their computer doing a bunch of background processing. You’re better off letting people rent ou…

>AWS at least already has very cheap ways to get compute time

No it does not. AWS and "very cheap" simply don't belong in the same sentence.

Re: Turning web traffic into a Super Computer

#16

Post by archgoon »

fwip wrote: post #9
AtomicOrbital wrote: post #2

opportunity here to create cloud computation platforms where people submit work which gets executed on other people's machines who are getting paid to allow their browser to contribute compute ... eventually price of AWS / google compute engine will drop ... win win

The big problem I see to be solved with p2p computation is trust. In the example, the problem to be solved is "Do any of these strings hash to a known value?" An adversary can accept the job, sleep for a second, and report "Nope" and collect their payment, without ever doing any work. In contrast, I trust that if I run code on AWS, it'll actually run the code I give it. These are solvable problems, but important to t…

A bigger issue than lying (and randomly failing) clients is the fact that you're farming out your users data to third parties. You might be willing to trust a single entity like Amazon, Microsoft, or Google, who have a reputation to maintain / can be sued (for potentially a lot); but if your farming out to essentially anonymous individuals, there's no repercussions for, say, using their credit card information.

Now, there are definitely use cases where you don't care; which would be a cool application for this; but unless we figure out how to do homomorphic computing, this doesn't seem immediately likely to impact cloud computing prices; even if it was cheaper to simply use excess compute capacity (which I'm not convinced it is).

Re: Turning web traffic into a Super Computer

#18

Post by cercatrova »

fwip wrote: post #9
AtomicOrbital wrote: post #2

opportunity here to create cloud computation platforms where people submit work which gets executed on other people's machines who are getting paid to allow their browser to contribute compute ... eventually price of AWS / google compute engine will drop ... win win

The big problem I see to be solved with p2p computation is trust. In the example, the problem to be solved is "Do any of these strings hash to a known value?" An adversary can accept the job, sleep for a second, and report "Nope" and collect their payment, without ever doing any work. In contrast, I trust that if I run code on AWS, it'll actually run the code I give it. These are solvable problems, but important to t…

This is what Golem cryptocurrency does, with the trust protocol in place through the blockchain [1]

1. https://golem.network

Re: Turning web traffic into a Super Computer

#19

Post by 0xFFFF0000 »

Something like this could be a way to get finally rid of advertisements, and help finance free services in a novel new way if the CPU isn't over committed. I thought the same around crypto mining... Many people might prefer the resource providing/computation option compared to ads.

Re: Turning web traffic into a Super Computer

#20

Post by noahdesu »

clon wrote: post #17

This will only work if you can cheaply validate the work performed. Or send it to several distinct visitors and compare results, driving down the likelyhood that the results are bogus.

good point. my hunch is that the only way would be to duplicate work and check for consensus. it seems like there isn't any way to cheaply validate results for any interesting work.
Post reply on HN