A few weeks ago Martin Sevior and Tom Ffield of the University of Melbourne did a talk at VPAC called “Belle Monte-Carlo production on the Amazon EC2 cloud” based on a paper they’d presented at the International Conference of Computing in High Energy and Nuclear Physics. The presentation is now available on the VPAC website.
It’s all about testing the cloud computing model via Amazon EC2 for Monte Carlo production for the SuperBelle experiment at the KEK collider in Japan. My favourite comment is that for a real full production run on Amazon EC2 to be useful it would need to be able to return data from S3 to the KEK collider at 600MB/s (~4.7Gb/s) sustained.
I don’t know what Amazon would say to that – well, apart from maybe “no”. 🙂
NB: This is the talk I mentioned in the comments on Joe Landman’s blog post called “Cloudy Issues“.
S3 definitely doesn’t have that kind of performance right now, agreed … but these days 5Gb/s is not a crazy amount of bandwidth – especially for high energy physics.
See also network status for the LHC Optical Private Network – the LHC isn’t even online but the seem to average over 5Gb/s outbound: http://lhcopn.web.cern.ch/lhcopn/lhcopn-interfaces.html
Even to Australia, effectively “the outback” of the LHC grid as my colleague puts it so well, we will sustain ~ 1Gb/s during running, and already peak at ~2.5Gb/s and this is happening today.
I’d also mention that Amazon procures transit from Level 3, who certainly don’t have issues with connectivity 😛
However it is important to note that to achieve this level of service, you can’t continue to be “just a guy with a credit card”. If you’re going to fork over millions of dollars, a more formal agreement should be reached 🙂
As an update, we’re already pushing the boundaries of what a normal EC2 user does – we had to ask for an increased instance limit for our recent 800 core run. It is kind of crazy when you just turn on something the size of Tango from your browser 😉