McKinsey Clears Air on Cloud Computing- Says Amazon is twice as expensive

[Ajay ohri]

McKinsey, that fine think tank of intellectuals recently dubbed cloud computing as not making sense -thus trying to throttle in its infancy a paradigm that could make companies across the world more competitive than they are today by helping cut costs precisely when they need it the most. The attempt to paint virtualization rather than remote computing is another attempt to cloud the air rather than clear the air on cloud computing. Most consulting companies would have pointed out industry affiliations and disclaimers on which companies they are representing or have represented.

Read other comments at the NYT article
http://bits.blogs.nytimes.com/2009/04/15/when-cloud-computing-doesnt-make-sense/

Its study uses Amazon.com’s Web service offering as the price of outsourced cloud computing, since its service is the best-known and it publishes its costs. On that basis, according to McKinsey, the total cost of the data center functions would be $366 a month per unit of computing output, compared with $150 a month for the conventional data center. “The industry has assumed the financial benefits of cloud computing and, in our view, that’s a faulty assumption,” said Will Forrest, a principal at McKinsey, who led the study.

http://www.decisionstats.com/2009/04/mckinsey-attacks-cloud-computing-having-no-sense/

My take on this is here-

Cloud computing will have lower costs as economies of scale kick in, as they did for nearly all technologies. McKinsey partners must be having a hard time to meet their annual bonuses if they have not factored this basic assumption in their cost projections. Cloud computing just converts this to a mass infrastructure from the present scenario where you pay annual licenses for software that you use for less than 60 % in a day, and hardware that you find obsolete in 3-4 years, which is off course gives accountants a reason to help you with depreciation and tax benefits. Rent a computer in the sky is simpler - and you would not need any consultant to help advise what configuration you need.

Mckinsey has deep touches with the outsourcing industry in India from their seminal paper in 1999, to their first concept Knowledge center that helped start it, to their alumni across the outsourcing sector which satisfy a mutual symbiotic relationship particularly in business research. Cloud computing actually help with virtual teams - no need for server farms, IT bureaucracies and Indian outsourcing can actually reduce a lot of costs along with American direct users. The intermediaries and consultants would be affected the most.

Indeed I am speaking on the Cloud Slam 09, precisely on how cloud computing can help lower the digital divide by giving high power computing to anyone having a thin shell laptop with a browser. Developing countries need access to HPC to better plan their resources and growth in an environmentally optimized manner.

www.decisionstats.com

[Jim Starkey]

I sorry you don't like their conclusions, but you would more effective
challenging their assumptions, data, or logic. Attacking the author
isn't good a way to advance an argument.

Your argument is that cloud computing will have lower costs when the
economies of scale kick in. What economy of scale are you expecting to
materialize that Amazon doesn't currently enjoy?

Your argument on software costs doesn't hold water. You are hoping to
reduce the amount you pay for software, which is understandable, but you
don't suggest why software vendors would be willing to accept less
revenue for their products. Until you can convince software companies
to give more for less (and become less profitable in the process), this
isn't going to happen.

It makes perfectly good sense for software companies to sell software as
a service with a pricing model that makes sense to them and their
customers. But if you think they're going to rent their software by the
hour in other people's clouds, I'm afraid you're in for a big
disappointment.

There is a general tendency on this list to see things as people wish
them to be rather than as they are. This is not the way good to make
progress.

> www.decisionstats.com <http://www.decisionstats.com>
>
>
> >


--
Jim Starkey
President, NimbusDB, Inc.
978 526-1376

[Ajay ohri]

Having sitten through umpteen glossy analysis by the same pedigree I have reason to belive the authors are biased in the analysis. Maybe it seems like I told you so, But I sat through six years of the M in action in offshoring and saw how they used to create the hype and then sell it. The McKinsey guys are pushing virtualization ahead of cloud computing but that is more of an attempt to hold back the tide. Any software vendor would move to cloud licensing because the hardware costs are lowered for the customer and they can be split up.
While the per hour cost of computing for that software will be higher -it will be change from the annual license model ( and I refer to data mining and predictive analytics )

Are you a virtualization chap - if so you may feel like cheering them on. One major vendor in predictive analytics and BI,

SAS has already commited to it's own private cloud. and with open source solutions in R like Biocep - this trend will accelerate.

You can comment here - instead of turning this into a huge back and forth match.

http://smartdatacollective.com/Home/17942

On Thu, Apr 16, 2009 at 8:10 PM, Jim Starkey <jsta...@nimbusdb.com> wrote:

I sorry you don't like their conclusions, but you would more effective
challenging their assumptions, data, or logic.  Attacking the author
isn't good a way to advance an argument.

Your argument is thaten cloud computing will have lower costs when the

[Kevin Apte]

I think McKinsey's conclusions are obvious from the Enterprise Architect @ Fortune 500 company side. 
The right answer I think is also quite obvious: It is to create private clouds for inhouse use, and make spare capacity available to others.
Eucalyptus project from UC Santa Barbara allows creation of  private clouds, that mimic Amazon's cloud offering.  Sun is allowing creation of Private Clouds. Microsoft gave hints that they may shift their position on creation of private clouds.

Amazon or SalesForce.com hosted software can be significantly cheaper on a per month basis, and not having high upfront expense is very
attractive to many companies.  Cloud Hosted Vertica for example, could be  a significantly cheaper  DataWarehousing tool than inhouse installed datawarehousing tools.

In general terms, and oversimplifying, I would say that the "Portal Layer"  could easily get cloud hosted. Applications with relatively few integration points could also get cloud hosted.
An application like SAP which probably has few hundred integration points- moving it to the cloud will be very difficult.  Many custom applications are simply too well entrenched in the IT landscape of a company, to be replaced by Cloud hosted equivalents.  

Cloud Computing is here to stay, and to grow rapidly.  It is the next big thing, like the Internet in the mid-90s. It is going to upset a lot of business models, and the industry landscape may not look the same 5 years from now. But every Cloud Computing company will not have the right marketable idea.

I am looking for people interested in creating a "Private Cloud" that meets the requirements for a Fortune 50 company.  This is the market that has not recieved much attention from anyone. I have some good ideas, and the bright minds here can only make them better. 

Kevin

[anthony...@gmail.com]

Completely agree!

Sent via BlackBerry by AT&T

---

From: Ajay ohri
Date: Thu, 16 Apr 2009 20:29:08 +0530
To: <cloud-c...@googlegroups.com>
Subject: [ Cloud Computing ] Re: McKinsey Clears Air on Cloud Computing- Says Amazon is twice as expensive

[Pietrasanta, Mark]

“I am looking for people interested in creating a "Private Cloud" that meets the requirements for a Fortune 50 company.  This is the market that has not recieved much attention from anyone. I have some good ideas, and the bright minds here can only make them better.”

Really?  This is what just about every hosting company is doing.  “Private Clouds” aren’t much more than virtualization (previously grid computing), when you get down to it, and you’d be hard pressed to find a hosting company that isn’t already doing this for small, medium, and large companies (and local, state, and federal government).

Amazon EC2 and the like can’t possibly survive given the current pricing models, and it’s refreshing to see a report finally say so.

[Larry Scott]

I spent time yesterday with the CIO and technology leadership teams at two US financial services firms.  While I will grant that some of the more cutting edge adopters of technology in the FS space are and have been users of cloud services, far too many have not.  These two were not.

I described to one leader how my firm was building a next generation data management platform in the cloud.  No physical servers, all open source, pay only for the server time that we utilize, deploy from the cloud to physical servers (hopefully virtualized, but not required) as and when needed.  When I indicated that a new server, pre-configured to our specifications, could be provisioned and made available in a matter of minutes, all he could do was laugh.  "What, no committee to review with, no approvals?"  

The lesson here is that there are vested interests in today's enterprise technology organizations who rightly view the cloud as competition.  I am sorry to say but it is the 40 somethings (and yes, I am in the category, reluctantly) who would seem to be the naysayers.

While I have a lot of respect for McKinsey, I would challenge them to show me how a PAAS model is more expensive than the traditional data center based server deployment.  Most CIOs and CTOs I know actually lament the cost of heat/power/space in the DC as the bigger challenge for them, not the box or rack itself.  I think rather than talk, we need more concrete testimonials of cloud computing deployments in the real world.  I'll be happy to offer up as we near completion.

Larry

--
Lawrence Scott
+1 617 834 0183

[Jim Starkey]

Fine. Let's assume they're biased. That doesn't mean they're wrong,
however. Humor us with your explanation of where they went wrong, if
you have one.

[Ajay ohri]

pricing of all technologies I have seen has decreased over time ( and I would like to see a single technology that didnt fall in price affordability). Technology adoption shows a S shaped curve. What the McKinsey model does is assume the Amazon price as static even though these are early days, and that price will fall as better competition, technologies , and scale would kick in.

This is like a consultant to ATT saying mobile phones are going to be too expensive based on the price of mobiles in 1984, but wireless handsets could be an option.

[Utpal Datta]

Do we all have our heads buried in sand and don't want to look at something just because the space is claimed sqarely by two "big" guys and there isn't much that is apparently obvious for the startups to do?

 

What do you think Azure is targeting? Even though MS wants people to run their application on MS cloud, Azure IS a platform to create Private Cloud.

 

And why there is no mention anywhere in this forum about VMWare's VDC-OS initiative.

 

Between VMWare and MS, 95+% of the virtualization market is taken. Also both the companies are in a strong position to provide storage virtualization (using the existing Data Center Storage hw). There you have it, a virtual pool of processors and a virtual pool of storage from which the enterprise (and small-med) users can provision on demand (and release resources when done) thus providing the "Elasticity" everyone here is talking about but cannot provide a practical solution.

 

Major benefits,

1. No data movement (in and out of Data Center) issues
2. No Security issues
3. Increase the current resource utilization from 15-20% to 60-70%
4. Full control of SLA is in the hands of Data Center Admins

Why would any user not take this obvious first step before venturing into something unknown, and unproven just because they make some good "science project"?

 

We also saw some comment that "Private Cloud" is not a cloud. Why?? (there I goi again into the vicious circle of "what is a cloud" :-))

 

I think the sooner we admit to the fact that "Private Cloud" is for real and there is less/little for the startups to do here, the sooner everyone will put their creative and innovative energy elsewhere, where startups have a good chance.

 

[Miha Ahronovitz]

Mark, 100% agreed.  McKinsey completely overlook the private cloud. Here are my calculations, based on McKinseyr supplied data:

McKinsey, which claims, based on Amazon.com cost studies, that data center traditional is more than half price the cloud computing alternative via Amazon .com ($150 per month per unit for data center vs $366 per month per unit for Amazon virtual cloud)

However, there is a couple of details here:

  1.Amazon.com is just a wholesale provider of resources in a monopoly position right now. Their prices will have to adjust when competitors are operational

  2. The price per year assumes someone will completely eliminate hardware acquisition and will replace with virtual resources from AWS (Amazon Web Services).

This is NOT the cloud  business case. The business case is to  supplement the during peak demand over a short period of time the physical resources from the locally operated private cloud.

In this case, we should consider the annual cost of the data center ($150 per unit per month of compute power) as being the exact cost of a private cloud. What we SHOULD compare, - assuming that we need double compute power for 1 month a year - the cost of buying incrementally hardware and using it only during peaks -  versus renting from Amazon for the times

Scenario A: we need more resources for a shorter time than 1 year, say a peak of 1 month duration.

  Buying: $150 *12 = $1,800 per year per unit of compute power, of which the utilization is 1/12 = 8%
  Renting from Amazon.com: $366 for 1 month only, 100% untilization

Savings   1,800 - 366 = 1,434 per year, or 80% cost savings.

Scenario  B: Assuming the peak is 2 month over 1 year:

Buying: $150 *12 = $1,800 per year per unit of compute power, of which the utilization is 2/12 = 16%
Renting from Amazon.com: $366 for 2 month only = 732 per year per unit of time 100% utilization

Savings   1,800 - 732 = 1,434 per year, or 59%  cost savings

from the above we can see, based on McKinsey cost data,  it makes sense to rent AWS resources for 1,800/366 = maximum 5 months in one year. More than that, we are better off to buy our own private cloud.

I think McKinsey should follow this group more carefully before they publicize their conclusions to un-informed large masses of readers.
Nevertheless, they did promote themselves with all the buzz and mis-information

2 cents,

Miha

---

From: "Pietrasanta, Mark" <Mark.Pie...@aquilent.com>
To: "cloud-c...@googlegroups.com" <cloud-c...@googlegroups.com>
Sent: Thursday, April 16, 2009 9:14:43 AM

Subject: [ Cloud Computing ] Re: McKinsey Clears Air on Cloud Computing- Says Amazon is twice as expensive

[eprpa...@gmail.com]

The pricing concept only works for mass produced products. Not all
technology decreases in price as time moves forward.

Chuck Wegrzyn
Twisted Storage

[Jim Starkey]

Reality check: Amazon has raised the price for four core instances from
$.70 to $.80 an hour.

So I guess you're faulting McKinsey for using Amazon's prices to
evaluate the cost of Amazon's service. Yup, definitely a sign of bias.

> > <mailto:jsta...@nimbusdb.com <mailto:jsta...@nimbusdb.com>>>

[Nik Simpson]

On 4/16/2009 1:13 PM, Miha Ahronovitz wrote:
> Mark, 100% agreed. McKinsey completely overlook the private cloud. Here
> are my calculations, based on McKinseyr supplied data:
>
> McKinsey, which claims, based on Amazon.com cost studies, that data
> center traditional is more than half price the cloud computing
> alternative via Amazon .com ($150 per month per unit for data center vs
> $366 per month per unit for Amazon virtual cloud)
>

I think you can make the case that McKinsey's recommendation to go for
server virtualization now isn't in conflict with this. The vast majority
of internal clouds will use virtual servers, so any work put into server
virtualization now will payback later when internal clouds are something
that they can implement.

--
Nik Simpson

[Jim Starkey]

Someone questioned the price change, and I haven't been able to verify
it. I suggest it be disregarded. The calculations I've reported at
various times were based on the $.80/hour. Most comparisons, of course,
are based on the 1 core, 32 bit VM, for which there is no equivalent
physical server still on the market, though an Atom based Netbook comes
close.

As the McKinsey study, give 'em hell for reporting the truth.

Jim Starkey wrote:
> Reality check: Amazon has raised the price for four core instances from
> $.70 to $.80 an hour.
>
> So I guess you're faulting McKinsey for using Amazon's prices to
> evaluate the cost of Amazon's service. Yup, definitely a sign of bias.
>
>
>

[Jan Klincewicz]

If I were Amazon, I'd hire McKinsey to do a buch of studies and write some expensive White Papers.

---

From: Miha Ahronovitz <mij...@sbcglobal.net>
Sent: Thursday, April 16, 2009 6:55 PM
To: cloud-c...@googlegroups.com
Subject: Re: McKinsey overlooked the Private Cloud Re: [ Cloud Computing ] Re: McKinsey Clears Air on Cloud Computing- Says Amazon is twice as expensive

After this McKinsey study, if I were Amazon, I would adjust the costs for 1 year commitments below $150 per node per compute power unit, down from $360. Even then the private cloud, will not be rendered as non competitive, because many problems (like HPC) can not be solved on an AWS virtual cloud.

What private clouds need is not virtualization for virtualization sake (Jim Starkey wrote  "you may wear out your teeth from gritting them every time you hear it.."). Private Clouds need elasticity on demand and virtuaization is one way to achieve this, among many.

The word  "elasticity" also implies that there is a business case for transient high loads, This requires have more resources readily available, no matter how many servers and storage the private cloud has. The calculation I did shows that we pay more for these supplementary peak resources on demand, As you say, "here are far cheaper ways to achieve that same end goal, no matter what use-case you want to throw at it"

miha

---

From: "Pietrasanta, Mark" <Mark.Pie...@aquilent.com>
To: "cloud-c...@googlegroups.com" <cloud-c...@googlegroups.com>

Sent: Thursday, April 16, 2009 2:24:33 PM
Subject: RE: McKinsey overlooked the Private Cloud Re: [ Clo

[The entire original message is not included]

[Kevin Apte]

---

From: "Pietrasanta, Mark" <Mark.Pie...@aquilent.com>
To: "cloud-c...@googlegroups.com" <cloud-c...@googlegroups.com>
Sent: Thursday, April 16, 2009 9:14:43 AM
Subject: [ Cloud Computing ] Re: McKinsey Clears Air on Cloud Computing- Says Amazon is twice as expensive

“I am looking for people interested in creating a "Private Cloud" that meets the requirements for a Fortune 50 company.  This is the market that has not recieved much attention from anyone. I have some good ideas, and the bright minds here can only make them better.”

Really?  This is what just about every hosting company is doing.  “Private Clouds” aren’t much more than virtualization (previously grid computing), when you get down to it, and you’d be hard pressed to find a hosting company that isn’t already doing this for small, medium, and large companies (and local, state, and federal government).

I am not aware of anyone who gives the complete IBM or Oracle SOA stack- Portal, BPM, Application Server, Data Service, Databases, Datawarehousing and analytics, Business Intelligence software as a well coordinated tuned package, at price far lower than inhouse prices.   There is also  a role for hosted SAP, PeopleSoft, J D Edwards, Oracle Applications and so on...

GigaSpaces has an impressive offering, for those applications that can convert to GigaSpaces. STAX also has a J2EE offering.   They are quite appropriate for new projects where the technology stack offered can be used.

Most hosting companies provide hosts, it is upto the client to install software, manage licenses and so on.  This is the model that McKinsey report talks about when pointing out that moving the DataCenter outside of the corporation offers very little benefits.

I am looking to privately talk to people who have worked in Fortune 500 companies. I will be happy to share my ideas.  My ideas are possibly a little unformed at this point, that is why I do not want to share them. But a hosted IT stack is not being offered as of now. It is too complex an undertaking for most hosting companies.

Kevin

[Ray Nugent]

McKinnsey gave this report at The Uptime Institute, who's membership is dominated by data center operators and Managed Service Providers, cloud computing's main competition...

The fact of the matter is the no consultant creates a report without being paid - sponsored - by someone (most likely the Uptime institute in this case). But the report does make good points about the hype and the mistaken use cases touted by cloud proponents. I remember hearing almost the same level of hype/promises in the mid 1990 about the World Wide Web. In fact, I once pitched an ecommerce proposal to a senior exec who told me point blank that no one would every buy our type of product over the internet. Two years latter the company was doing $2 billion in business via the web.

All of the claims and counter claims are meaningless at this point. Prices will adjust and things like security will get solved or proven to not be problems after all. But for now the debate will probably continue to rage until there are anough practitioners to make debat irrelevant.

Ray

---

From: Kevin Apte <kevi...@gmail.com>
To: cloud-c...@googlegroups.com
Sent: Thursday, April 16, 2009 6:20:51 PM
Subject: Re: McKinsey overlooked the Private Cloud Re: [ Cloud Computing ] Re: McKinsey Clears Air on Cloud Computing- Says Amazon is twice as expensive

[Rao Dronamraju]

I am not sure how many have seen the actual presentation….here it is….

 

http://uptimeinstitute.org/images/stories/McKinsey_Report_Cloud_Computing/mckinsey_clearing_the%20clouds_final_04142009.ppt.pdf

 

Their case based on ONE public cloud provider is almost laughable. The only thing I agree with them is the 20+ definitions of clouds.

 

 

 

 

[Rao Dronamraju]

"Your argument is that cloud computing will have lower costs when the
economies of scale kick in. What economy of scale are you expecting to
materialize that Amazon doesn't currently enjoy?"

Well, how much of the SME industry, just in US alone, is now in the public
clouds?...0.00001%?...You call that economy of scale?....Now think about SME
across the world?...

"Your argument on software costs doesn't hold water. You are hoping to
reduce the amount you pay for software, which is understandable, but you
don't suggest why software vendors would be willing to accept less
revenue for their products. Until you can convince software companies
to give more for less (and become less profitable in the process), this
isn't going to happen."

I think it is a completely new pricing/revenue model that has never been
tried before. Before Internet/Web, how many companies made money through
ONLY ADVERTISING?...Haven't most internet/web companies thrived under this
new revenue/pricing model?...

I think the biggest problem with McKinsey report/study is their entire case
is based on ONE vendor. In addition, the whole cloud industry is just
forming. You never know how many different forms and shapes it can take
before it gets to a quiescent state, probably atleast 5 years during which
time there will rapid price declines and stablization as happened in every
other industry that had a highly disruptive affect.

I think we all can look back at the web/internet markets/industry and learn
from it. Who thought giving away internet services for FREE would flourish
SO WELL for SO LONG. Even today there isn't any talk about moving away from
the FREE model of internet.

So I think McKinsey's report is just too damn premature.

-----Original Message-----
From: cloud-c...@googlegroups.com
[mailto:cloud-c...@googlegroups.com] On Behalf Of Jim Starkey
Sent: Thursday, April 16, 2009 9:40 AM
To: cloud-c...@googlegroups.com
Subject: [ Cloud Computing ] Re: McKinsey Clears Air on Cloud Computing-
Says Amazon is twice as expensive

[Pietrasanta, Mark]

Okay, that clarifies that you’re looking for a SaaS/hosted application offering, not simply a private cloud.  (Of course, everything and anything is apparently a cloud nowadays…)

[Miha Ahronovitz]

Look at slide 24:  It assumes typical CPU/month cost for 3GHz dual-core Xeon Windows-based server. Sure this cost is lower. Then it assumes the 100% of the data center is moved to AWS. This is not something  to do.

The Private Clouds do appear on slide 30. McKinsey dismisses them as “Publicly-announced private clouds are essentially an aggressive virtualization program on top of the traditional enterprise IT stack”

 

If they did not like any cloud definition, why they don’t make one definition themselves?

And if they have no cloud definition at all  to work with, what do they analyze? Something they don’t know if it is a cloud or not?

 

Miha

 

No virus found in this incoming message.
Checked by AVG - www.avg.com
Version: 8.5.287 / Virus Database: 270.11.58/2062 - Release Date: 04/16/09 16:38:00

[Pietrasanta, Mark]

Perhaps I missed it, but where did they announce EC2, non-advertising supported, non-S3/AWS revenues?

And in any case, given their pricing, they better be doing okay. Just because something makes no sense doesn't mean people won't use it. There are endless examples of overpriced services that thrive for short periods of time.

---

From: Ray Nugent <rnu...@yahoo.com>
Sent: Friday, April 17, 2009 9:55 AM
To: cloud-c...@googlegroups.com <cloud-c...@googlegroups.com>
Subject: Re: McKinsey overlooked the Private Cloud Re: [ Cloud Computing ] Re: McKinsey Clears Air on Cloud Computing- Says Amazon is twice as expensive

Again, I don't think you can deny facts Mark. Regardless of what you consider to be expensive, they seem to be doing a pretty good business. If you have some evidence to the contrary, please provide it.

---

From: "Pietrasanta, Mark" <Mark.Pie...@aquilent.com>
To: "cloud-c...@googlegroups.com" <cloud-c...@googlegroups.com>

Sent: Thursday, April 16, 2009 10:10:43 PM

Subject: RE: McKinsey overlooked the Private Cloud Re: [ Cloud Computing ] Re: McKinsey Clears Air on Cloud Computing- Says Amazon is twice as expensive

No, they’re more expensive than any other model you can put together, and they have no SLA to speak of.

 

From: cloud-c...@googlegroups.com [mailto:cloud-c...@googlegroups.com] On Behalf Of Ray Nugent

Sent: Thursday, April 16, 2009 8:08 PM
To: cloud-c...@googlegroups.com

Subject: Re: McKinsey overlooked the Private Cloud Re: [ Cloud Computing ] Re: McKinsey Clears Air on Cloud Computing- Says Amazon is twice as expensive

 

The fact of the matter is they have a thriving business because they provide use cases that make economic sense for a certain market. Your argument has been consistently "they're more expensive than colo" and, in one use case, your right. CPU to CPU costs. However, the intrinsic value of cloud computing is "on demand". This is not unlike the trends in manufacturing in the 1980's that emphasised just in time material delivery. Why take posetion of something and pay for it while it sits idle? That's what you do in 'every other hosting model". So, if one was to move an entire enterprise dataceneter into the cloud, CPU for CPU, that would certainly cost more. However, if one where to do so and leverage an on demand useage modle, eliminating over provisioning and reducimg costs to deploy and surrender machines their costs would drop considerably from a traditional physical datacenetr model.

Ray

 

---

From: "Pietrasanta, Mark" <Mark.Pie...@aquilent.com>
To: "cloud-c...@googlegroups.com" <cloud-c...@googlegroups.com>
Sent: Thursday, April 16, 2009 4:00:58 PM
Subject: RE: McKinsey overlooked the Private Cloud Re: [ Cloud Computing ] Re: McKinsey Clears Air on Cloud Computing- Says Amazon is twice as expensive

It has nothing to do with “wishing”.  It has to do with them charging multiple times what any other hosting model costs.  There is no way that can survive.  Either they change their pricing model, or they fade away.  Simple economics.

 

From: cloud-c...@googlegroups.com [mailto:cloud-c...@googlegroups.com] On Behalf Of Ray Nugent

Sent: Thursday, April 16, 2009 5:43 PM
To: cloud-c...@googlegroups.com

Subject: Re: McKinsey overlooked the Private Cloud Re: [ Cloud Computing ] Re: McKinsey Clears Air on Cloud Computing- Says Amazon is twice as expensive

 

That's a pretty bold statement given they have been more than surviving for almost three years. Just because you wish they would fail doesn't mean they will. - Ray

"Amazon EC2 and the like can’t possibly survive given the current pricing models, and it’s refreshing to see a report finally say so."

From: "Pietrasanta, Mark" <Mark.Pie...@aquilent.com>
To: "cloud-c...@googlegroups.com" <cloud-c...@googlegroups.com>
Sent: Thursday, April 16, 2009 2:24:33 PM
Subject: RE: McKinsey overlooked the Private Cloud Re: [ Cloud Computing ] Re: McKinsey Clears Air on Cloud Computing- Says Amazon is twice as expensive

“Mark, 100% agreed.  McKinsey completely overlook the private cloud. Here are my calculations, based on McKinseyr supplied data:

McKinsey, which claims, based on Amazon.com cost studies, that data center traditional is more than half price the cloud computing alternative via Amazon .com ($150 per month per unit for data center vs $366 per month per unit for Amazon virtual cloud)

However, there is a couple of details here: <snip>”

Except that ignores the “private cloud” hosting facility.  They have similar abilities to offer short-term additional capacity.  And they typically do it at a fraction of the cost of Amazon EC2.  Not in the exact same manner, but it’s close enough.  And it’s even better if you’re hosting with them to begin with.

I don’t own, sell, or promote any hosting business.  Frankly, I find most of them to be simply (often annoying) vendors selling commoditized solutions.  But the reality is, many/most of them have moved to virtualization, and many of them have some sort of shared offering.  While it’s likely not at the same scale as EC2, it certainly is good enough for most people under most circumstances, and the pricing tends to be much more in-line with what you’d expect/want.

I guess I don’t see any additional value in something like Amazon EC2 over smaller private virtualized environments.  I think “cloud computing” as a whole will go beyond simply “perceived infinite capacity with instant provisioning” virtualized environments, but right now that’s all they seem to be, including EC2. 

And there are far cheaper ways to achieve that same end goal, no matter what use-case you want to throw at it.

 

---

From: "Pietrasanta, Mark" <Mark.Pie...@aquilent.com>
To: "cloud-c...@googlegroups.com" <cloud-c...@googlegroups.com>
Sent: Thursday, April 16, 2009 9:14:43 AM
Subject: [ Cloud Computing ] Re: McKinsey Clears Air on Cloud Computing- Says Amazon is twice as expensive

“I am looking for people interested in creating a "Private Cloud" that meets the requirements for a Fortune 50 company.  This is the market that has not recieved much attention from anyone. I have some good ideas, and the bright minds here can only make them better.”

Really?  This is what just about every hosting company is doing.  “Private Clouds” aren’t much more than virtualization (previously grid computing), when you get down to it, and you’d be hard pressed to find a hosting company that isn’t already doing this for small, medium, and large companies (and local, state, and federal government).

Amazon EC2 and the like can’t possibly survive given the current pricing models, and it’s refreshing to see a report finally say so.

On Thu, Apr 16, 2009 at 10:40 AM, Jim Starkey <jsta...@nimbusdb.com> wrote:

[Randall Minter]

>If they did not like any cloud definition, why they don't make one
>definition themselves?

They did, their definition is:

Clouds are hardware-based services offering compute, network and storage capacity where:
▪ Hardware management is highly abstracted from the buyer
▪ Buyers incur infrastructure costs as variable OPEX
▪ Infrastructure capacity is highly elastic (up or down)

- Randall

[jgb...@gmail.com]

On the face of it, I don't disagree with the McKinsey report, or even
with the "sound bite" conclusion that outsourcing a complete (large
enterprise) data center to a cloud provider does not make economic
sense because the costs are 2X higher. There is a lot of truth to this
report, even if you want to quibble with the details. A key aspect of
clouds, from the purchaser's point of view, is that it moves the costs
from the capital side to the operational side, with a pay-per-use
model and little or no upfront costs or contractual lock-in. This,
plus the elasticity of dynamic scale-up / scale down, is the key value
proposition for small and medium business, especially startups. The
McKinsey report does not dispute this, it in fact validates it.

Expecting the same value proposition to apply to the enterprise is not
realistic. If it were, nobody would actually buy a car (use in-house
data center), or lease a car (traditional hosting), we would all rent
them by the day (cloud computing). In fact, only a 2X cost increase to
outsource a whole corporate data center to the cloud is a very
encouraging number -- I think it is surprisingly low! Consider again
the buy-a-car vs rent-a-car analogy. Renting a car by the day or week
is on order 4X more expensive than buying one.

So I agree with large parts of the McKinsey report, but I think in its
conclusions to some degree it throws out the baby with the bath water.
It assumes that the question "Is cloud computing useful in the
enterprise?" is equivalent to "Does it make sense to outsource the
whole enterprise data center to the cloud?". Certainly, cloud
computing is useful to the enterprise now, and certainly it needs more
work (better security, true SLAs, more interoperability, etc.) to be
more even useful in the future, but no, it doesn't make sense to
outsource the whole enterprise to the cloud now. Cloud is useful, but
it has a different value proposition to the enterprise. In my view,
quick scale up / scale down for extra capacity spikes without capital
outlay is the key value prop for enterprise use of external clouds.

I also strongly disagree with a key second-level conclusion in the
McKinsey report -- that CIOs should dismiss "internal clouds" and just
concentrate on virtualization. Clearly, using virtualization to get
average utilization rates up north of 10% is valuable in and of
itself. But that is not the whole story. "Internal clouds" make
abundant sense in the enterprise, and many exist or are being
implemented today. Cloud computing's instant start-up model without up
front capital investment and delays applies internally in the
enterprise as well.

In the typical traditional enterprise today, the IT end user /
consumer often does not get what they need when they need it. Most
enterprise IT shops operate on some chargeback model. The IT
organization may hold title to the server, or the using organization
may hold title, but if you need a server resource, *somebody* has got
to justify and acquire a server, often through a cumbersome capital
acquisition process, then provision it, install the application stack,
and maintain it. This all takes far too much time in the traditional
enterprise environment -- days or weeks or even months. Sometimes
their (unanticipated) needs are out of sync with the corporate funding
cycle, or they have the wrong flavor of money (e.g., they have
operating expense money, but don't have capital, or vice versa), but
one way or another, it almost always just takes too long to get what
you need. As a result, the business can't respond quickly to
opportunities or problems. An internal cloud can help.

With an internal cloud, implemented well, this friction and lost work
can largely go away. Set up a bunch of servers to be managed as a pool
(or grid) of virtual resources, with a front end to allow self-service
requests or scheduling, automation to do the repetitive stuff, a
reasonable chargeback model, and some policy to help control sprawl,
and you have an "internal cloud". Need a resource now, get a resource
now, and get charged for what you use.

The speed (or elasticity) of the cloud -- the ability to get or change
an IT resource quickly -- is extremely valuable internally in the
enterprise, but the McKinsey report ignores this aspect completely by
focusing on only on costs and utilization.

-- Joe Baron

[Ajay ohri]

some of the mckinsey conclusions are startling enough- wonder what Amazon has to say. Surely the expensive part must have hit a nerve.

read more here

http://smartdatacollective.com/Home/17942

[Michael Grove]

Personally I think the definition, and top down vs. bottom's up convergence framework makes sense. What is not well discussed are the key drivers. Greater variable/fixed costs, SAAS plug and play, and IT self-service will be key factors to roadmap decisions in my view. .IT needs to be pervasive, highly agile, and yet well managed. Its all about making money and being better. The cloud fits into a list of strategic options, not the other way around. Mike Michael Grove 650-346-8059(M) michae...@collabworks.com www.collabworks.com --- On Fri, 4/17/09, jgb...@gmail.com <jgb...@gmail.com> wrote:

From: jgb...@gmail.com <jgb...@gmail.com> Subject: [ Cloud Computing ] Re: McKinsey Clears Air on Cloud Computing- Says Amazon is twice as expensive

[Greg Pfister]

On Apr 16, 5:47 am, Ajay ohri <ohri2...@gmail.com> wrote:
> McKinsey, that fine think tank of intellectuals recently dubbed cloud
> computing as not making sense -thus trying to throttle in its infancy a
> paradigm that could make companies across the world more competitive than
> they are today by helping cut costs precisely when they need it the most.
> The attempt to paint virtualization rather than remote computing is another
> attempt to cloud the air rather than clear the air on cloud computing.

The problem here is not that they found one case where it doesn't make
sense -- nothing makes sense for *everything* -- but that everybody's
overgeneralizing that to all cases. Even the NYT blog entry qualified
the statement.

Of _course_ if you already run a large data center yourself with state-
of-the-art efficiency you aren't going to win by paying someone else
to do it. Above some size, there aren't significant additional
economies of scale.

The actual slides do say "Clouds already make sense for many small and
medium-size businesses" but not large enterprises.

I agree McKinsey plays down private clouds too much -- or maybe
they're distinguishing private clouds and private grids, which are
already in regular use.

This whole discussion is for me a tempest over nothing particularly
surprising or even interesting.

Greg Pfister
http://perilsofparallel.blogspot.com/