Real Time Services: Pricing models and Grid computing

If we see the history, in 19th century there were two utilities that changed our life ie: electric and telephone. As Grids already proved in simulations, exploring universe, etc. and I believe on demand access to computing, data and services would become part of our daily life, very similar to electricity, gas, water etc. Even though Grid computing is in so early stage, efforts of open standards providing strong hopes for mass adopting. In this series of posts I am analyzing about 'pricing models' of various commercial Grids available today, started with some introductions…

People talking about ‘Software as a service’ more than a decade (I first heard this term from Bill Gates key note in 2000), but I don’t think it has got big success till this date, the trend is slowly changing, particularly when there is a need for ‘Real Time digital Service to people’ and ‘enough infrastructure (Grid) in place’, this where SaaS going to get great success,
“Success of real time service = Collaborative Enterprise + Shared Infrastructure (Grid) + Shared Software Services”

Previously I said, Grid is a utility model, that doesn’t mean it can be compared as it is with Telephone/Electricity since which are not ‘shared’ moreover shared infrastructure is creating eco-system by default.

Grid pricing is based on all the factors that I listed in the formula, First I like to rant about ‘Shared Infrastructure’, below are few recognized models…

The most dominant mode is remote hosting; this model is more closely a replication of ‘Web Hosting’ which is suitable for Small to Medium business (ie: tail in the Long Tail). The second model is ‘In-house hosting’ which managed by remotely or by the Enterprise itself (ie: Heads in Long tail) , The third model, ‘Personal Hosting’ which is special remote hosting, where end users directly host their personal stuff (ie: personal bots/crawlers, Games, rendering etc) in Grid infrastructure, which is suitable for generally Geeks, Media professionals or 24/7 web users J etc

Few factor which makes pricing more complex...
- There is no standard followed in pricing model, each vendor follows their own calculations based on different usage scenarios, its bad for customers, how they can compare price and select among the vendors?
- Software license model typically follows per-seat or per CPU basis; this is not suitable for utility pricing. I believe that if the current licensing model changed, there would be a heavy competition among software vendors, and customers could change the vendors almost instantly, it’s also creating instability in projecting future revenue stream for software vendors, Maximum one/two months can be clearly projected!
 
When user sends request to Grid for new/dispose resources, handshaking happens between user request and grid infrastructure to determine what resources the user gets from the pool, then on, the infrastructure keeps recording the usage to track resource usage for billing.
 
"Choosing hosting model depends on business model"
Few of my observations that would help to identify the hosting model...
- If utility infrastructure is outsourced, the pricing model more closely a fixed price, few may not agree that this is not utility computing but still this model exist with few vendors. Fixed price is suitable for applications which serve 24/7. If multiple applications means may need lot of calculation to choose wisely.
- Hybrid model charges a fixed basic price and per-use price,

- Pure per-use pricing, this is considered standard best model

I would write about SaaS and few commercial Grid offering in future posts…

See Also:
Long tail - Grid computing is suitable for Head or Tail ? Part II
Long tail - Grid computing is suitable for Head or Tail ?

Post your comment