[Cloud Computing ] Billing the Cloud User
ken_...@compuserve.com
> You can only maintain usage pricing when you are selling something for
> which there is no substitute product - music and books being a good
> example. If I like Joe's particular kinds of guitar riffs, I can't
> just substitute Susan's music because it is cheaper. To the extent that
> vendors add unique value added services they can get away with things
> like transaction pricing. But if they selling generic DB IO or
> computing power then cost, not utility will drive prices.
In the late-60s and 1970s, there was a computer timesharing industry that's
the closest parallel to cloud computing as an industry. At its peak there
were dozens of vendors, including IBM, GE and Computer Sciences
Corporation. It was the mainframe era when a medium size computer leased
for $15-17,000 per month and a large one sold for $3+ million.
Organizations opened timesharing accounts for the same reason they'll open
an account with a cloud computing vendor - access to computing resources on
a pay per use basis, without having to lay out millions for a computing
infrastructure. This model works quite well if you are developing
applications or services on a project or contract basis.
Organizations also used timesharing accounts for overflow situations, such
as when they needed extra capacity for year end processing, long-running
simulations, reservoir modeling or other one off requirements. That's a
class of user who'll look to cloud computing today.
Timesharing vendors initially competed on the basis of price, uptime, queue
delays and operating system. If you were writing an app that was eventually
to be hosted on a Honeywell GCOS system, you selected a service such as
GE's GEISCO. If the target was Univac, you'd like use CSC Infonet. Cloud
computing removes platform dependency and scalability (queue delays) from
the picture, but price and uptime are still competitive factors.
As the timesharing space became crowded, vendors broadened the portfolio of
software available to their customers. They offered multiple applications
and DBMSs (also billed per usage) and supported their use with training and
tech support. That's likely to be a differentiator today. Look at what
AppExchange has done for salesforce.com.
======== Ken North ===========
www.KNComputing.com
www.DataServicesWorld.com
www.CloudComputingSummit.com
www.LinkedDataSummit.com
www.WebServicesSummit.com
www.SQLSummit.com
www.GridSummit.com
--------------------------------------------------------------------
mail2web - Check your email from the web at
http://link.mail2web.com/mail2web
JM
timesharing will be more than interested in cloud computing today.
Also I believe an ecosystem of services (composite applications)
within the cloud will be interesting as well if there is a feasible
charge back method. Here is a blog post I put together, I think new
start-ups today will be able to monetize on shared services better
than what we have seen since late '90s. There are alot of SaaS apps
out there for free or on a monthly basis. For transactions, processes
or requests triggered by a human a monthly fee might work well
regardless of the usage just because of the available house in the day
and times you can hit a mouse button. On the other hand system-to-
system transaction typically found in the application management,
systems management, etc space are dependent on actual elapsed CPU time
usage for chargebacks. If one company processes a composite
transaction 1 Billion times a month the fee should be different than
if a company uses 1 million composite application transactions.
http://www.camsolutionsinc.com/Blog/bid/7347/Business-Process-Management-BPM-Cloud
Just my 2 cents.
JM
On Nov 14, 12:34 pm, "ken_no...@compuserve.com"
Jim Starkey
> Elizabeth Grace Frank-Backman wrote:
>
>> You can only maintain usage pricing when you are selling something for
>> which there is no substitute product - music and books being a good
>> example. If I like Joe's particular kinds of guitar riffs, I can't
>> just substitute Susan's music because it is cheaper. To the extent that
>> vendors add unique value added services they can get away with things
>> like transaction pricing. But if they selling generic DB IO or
>> computing power then cost, not utility will drive prices.
>>
>
> In the late-60s and 1970s, there was a computer timesharing industry that's
> the closest parallel to cloud computing as an industry. At its peak there
> were dozens of vendors, including IBM, GE and Computer Sciences
> Corporation. It was the mainframe era when a medium size computer leased
> for $15-17,000 per month and a large one sold for $3+ million.
>
> Organizations opened timesharing accounts for the same reason they'll open
> an account with a cloud computing vendor - access to computing resources on
> a pay per use basis, without having to lay out millions for a computing
> infrastructure. This model works quite well if you are developing
> applications or services on a project or contract basis.
>
>
May I point out that time shared died -- extinct-- very shortly after
VAXen and other suitable departmental computing platforms became
available. Timesharing was attractive only when there was no
alternative. The question was one of control, not price -- An entry
level 780 was way beyond anyone's timesharing bill.
Why is history going to reverse itself?
Ray Nugent
Ray
From: Jim Starkey <jsta...@nimbusdb.com>
To: cloud-c...@googlegroups.com
Sent: Saturday, November 15, 2008 11:58:54 AM
Subject: [ Cloud Computing ] Re: [Cloud Computing ] Billing the Cloud User
JM
infrastructure-as-a-service back then regardless of the hardware, and
cloud computing / infrastructure-as-a-service today you are
essentially talking about the same thing. $0.10 per hour on EC2 is
time sharing.
On Nov 15, 11:58 am, Jim Starkey <jstar...@nimbusdb.com> wrote:
Paul Renaud
that states that the price of an undifferentiated product will inevitably
become equal to its marginal cost. (This one of the insights that won Nash
his Nobel prize.)
In other words, it is a mathematical certainty that commodity cloud
computing will ultimately be priced at cost unless it is differentiated by
value-added services. That future world will very much resemble the old
timesharing world where economies of scale dominated because they had the
lowest marginal cost.
That future outcome does not have to be realized. Fortunately there are
many ways to add value.
Cheers, paul