Hi, regarding computational service as a market, I think it does not fit on any end user activity. But on the other hand the companies providing services to the end users may require some computational services.

For instance:

There is a service provide in the us that allows user to play games at home without any additional hardware besides a TV, Game-pad and internet connection (OnLive.com) . As games are very computational intensive application they need to a certain quality of service regarding computational power and they don't need all that power all the time. So I think that this example fits on your target market.

Amazon EC2 is a kind of service that delivers on demand computational power, but lacks on quality of service for services like OnLive, on the otherhand for a web service that needs to respond to the instantly increased number of requests if fits perfectly.

Although there are some challenges regarding this kind business, the most important in my opinion is: quality of service like desired latency between the request and the response.

To conclude, Yes I think there is a marked for computational services, not for end users but for the companies that provide services to end users and need some extra "juice" to comply with unexpected demands of service or to outsource those resources. The problem is that it may need to be a customer to customer tailored service.

Why not? have we imagined the services what we enjoy right now ,say, a decade ago?

But can such a market be extended from a purely scientific applicability to a real solution for business domain ?

Hi, regarding computational service as a market, I think it does not fit on any end user activity. But on the other hand the companies providing services to the end users may require some computational services.

For instance:

There is a service provide in the us that allows user to play games at home without any additional hardware besides a TV, Game-pad and internet connection (OnLive.com) . As games are very computational intensive application they need to a certain quality of service regarding computational power and they don't need all that power all the time. So I think that this example fits on your target market.

Amazon EC2 is a kind of service that delivers on demand computational power, but lacks on quality of service for services like OnLive, on the otherhand for a web service that needs to respond to the instantly increased number of requests if fits perfectly.

Although there are some challenges regarding this kind business, the most important in my opinion is: quality of service like desired latency between the request and the response.

To conclude, Yes I think there is a marked for computational services, not for end users but for the companies that provide services to end users and need some extra "juice" to comply with unexpected demands of service or to outsource those resources. The problem is that it may need to be a customer to customer tailored service.

This is possible and it is something that I'm working on right now. However, there are few challenges that need to be solved before such market evolve. For example, how to specify cloud resources/services? how to specify cloud user's requirements? how to map the user's requirements onto cloud resources.

Have a look at the following papers.

A commodity-focused multi-cloud marketplace exemplar application

http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6008759&tag=1

A constraints-based resource discovery model for multi-provider cloud environments

http://www.journalofcloudcomputing.com/content/1/1/6/

Specifying cloud application requirements: an ontological approach

http://www.chinacloud.cn/upload/2012-11/12110409165965.pdf

Commenting of the suggestion received from Yih Sun, I could see here at least one immediate solution for enhancing the operations within the emergent service market which is using Service Level Agreements. SLAs have a number of advantages when referring to user requirements and quality of services and they can be particularly useful for mapping service operations.

Here are the studies where we investigate this problem:

SLAs as a Complementary Currency In Peer-to-Peer Markets:

http://dl.acm.org/citation.cfm?id=1884562

Trading SLAs within a Peer-to-Peer Market

http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=5697974&tag=1

Going back to Francisco Correia's comment, the market can be initially applicable only for companies mapped as service providers and clients. But service providers often prefer to outsource the actual enactment/execution to a data centre –with the emergence of infrastructure providers such as Amazon.com and when this happens the latency of requests/responses becomes more significant.

We also explore this within 'Risk Assessment in Service Provider Communities ':

http://link.springer.com/chapter/10.1007/978-3-642-28675-9_10?null

Now a real challenge that I see in this context is to define a functional community of providers and clients with the associated trust relationships for enabling the exchanges of services.

From a business perspective I think developing multi-vendor cloud computing markets makes sense. Think about the multi-facetted configuration possibilities companies and private users of cloud computing services can choose from in areas such as SaaS, IaaS and STaaS.

In fact, due to the high level of intransparency on the cloud market independent providers segmenting cloud offerings in a sensible manner would offer immensley high values for end-customers. This would include intelligent comparison, rating and benchmarking capabilities.

While IaaS and STaaS cloud services could be compared on a quantiative basis (SLAs, metrics etc.), SaaS cloud computing services are harder to compare because they have so many qualitative characteristics (look and feel, integration etc.).

This is a list of current cloud computing markets:

www.getapp.com

appexchange.salesforce.com

www.cloudorado.com

Kind regards

Karl

Hi everybody,

We took the idea forward and developed a cloud computing marketplace for German SMEs called www.clouneo.com. Please have a look at it.

Kind regards

Karl