Lawyer Bait

The views expressed herein solely represent the author’s personal views and opinions and not of anyone else - person or organization.

Thursday, March 26, 2009

Podapalooza - Good time in Boston

The other night in Boston, I hosted a cocktail reception and dinner for local folks to come to the Westin in Boston and check out the HP POD, enjoy me buying them drinks and pulling together a lobster dinner/clambake. All of it amounted to a GREAT time.

I also wanted to do a little shameless self promotion on my blog, because I am starting to see the shift from pod computing being a WTF? solution to an interesting solution and people are starting to truly grasp how and why they make sense. A couple of examples from discussions over lobsters:

I could see using this to expand into a data center while my new data center is being built. I can use the PODs for two years and then drive them over to the data center when they have served their short term purpose and now I have a backup solution that I can put anywhere my risk managers want.

For genome sequencing this is ideal. We are no longer constrained by scheduling jobs. I know what I have for CPU cycles, memory, everything - and I can just run job after job after job. It will be the workhorse our facilities guys always gave us flack about building becuase we couldn't cool it.

I think when companies start spending again that the pent up demand will be such that if anyone has to get something up and running in a quarter - these PODS will be the best game in town. Imagine - a data center racked, stacked, and tested - delivered in 8 weeks. Wow. Just wow.

We will try to do more of these events in the near future. They are valuable, small, no spin, and FUN!





CRG WEST ANNOUNCES HP POD COMPATIBLE DATA CENTERS

CRG West Makes Deploying, Powering, and Securing HP Performance-Optimized Data Centers Easy

DENVER, CO– March 25, 2009 – CRG West is pleased to announce compatibility at multiple data centers across the United States with HP POD (Performance-Optimized Datacenter), a container-based data center offering. CRG West data centers in Boston, Los Angeles, and the San Francisco Bay Area can monitor, power, and secure HP PODs with unmatched efficiency.

HP PODs are a flexible, expedited solution for companies looking to augment existing data center infrastructure or deploy a ready-to-use data center IT solution. They support a wide variety of HP and third-party technology, increased power density supporting 3,500+ compute nodes or 12,000 large-form-factor hard drives and flexible configurations to meet specific customer requirements. In addition, HP offers supporting infrastructure services, such as assessment, preparation and deployment services, as well as data center design and planning. Each HP POD comes ready to use in a 40-foot container and requires approximately one megawatt to power.

Customers facing infrastructure, space, or power supply constraints can take advantage of this innovative technology by deploying their HP POD in a CRG West HP POD compatible data center. CRG West provides more than 2,000,000 square feet of robust, carrier-neutral data center space and 150+ megawatts of power across the United States. HP POD compatible data centers offer tailor-made deployment capabilities, with increased ceiling height, ample power and water supply, and crane-lift capable space, allowing for a deployment process that is as easy as “deliver, position, and power on.”

In doing so, customers can also take advantage of everyday CRG West amenities such as usage-based power pricing; 24-hour remote hands and security; redundant emergency generator power; access to 200+ carriers and ISPs; and a convenient on-line customer portal that allows for real-time power use tracking from any computer with an internet connection. In addition, utilizing CRG West data centers can eliminate the possibility of an outside deployment, where exposure to weather and other elements can require additional planning.

“CRG West HP POD compatible data centers make it easy for any prospective buyer to take advantage of the many perks, without jumping through power supply or space constraint hurdles,” said Mark Mac Auley, CRG West Vice President of Strategic Accounts. “These locations make an HP POD deployment as easy as ready-position-power, in a highly secure, protected CRG West data center environment.”

“Customers can maintain a competitive edge in the marketplace without having to invest in another building in order to build out their infrastructure,” said Steve Cumings, director of infrastructure, Scalable Computing and Infrastructure organization at HP. “The HP POD combined with CRG West’s nationwide, turn-key data center capabilities gives customers an easier way to deploy IT resources while minimizing expenses and improving data center efficiency.”

More information about the HP POD products, software and services is available at www.hp.com/products/pod.

An HP POD virtual video tour is available at www.hp.com/go/pod.

About CRG West:

CRG West is a leading developer, manager and operator of world-class data centers. Established in 2001, CRG West provides wholesale data center space and colocation, connectivity services, remote hands support and a public Internet peering exchange, the Any2 Exchange. A wholly-owned portfolio company and operating partner of The Carlyle Group, a global private equity firm with over $91 billion of equity under management, CRG West manages carrier-neutral data centers in Boston, Chicago, Los Angeles, Miami, New York, Northern Virginia, the San Francisco Bay Area, and Washington D.C. CRG West provides data center and peering opportunities to more than 500 of the world’s leading networks, enterprises, government institutions and universities. Visit www.crgwest.com for additional information.

Wednesday, March 18, 2009

It's all about Cosmos Computing

Clouds are too small for what is possible in the new computing paradigm.

Clouds are more than a Visio object or a piece of Powerpoint Clip Art.

Cloud up until now has been about defining what we know based on old models that are known. Grid, Utility, etc. Where we are going and what we are doing is larger than that - it is at the Cosmos level.

We can define the components but not the end state. We don't know what we
don't know but we understand that this is bigger than the sum of its known parts.

Clouds work because the Cosmos works

The Cosmos players are bringing the open structures, protocols, and relationships together that allow the components of Cloud computing to work.

Cloud computing has many interpretations, the Cosmos doesn't. It is. It works. It has worked for years and is now evolving. Again.

The Cosmos is elastic, resilient, interconnected, open, limitless, expanding, and alive.

The Cosmos binds the defined and undefined together into a system that is bigger than all of us, comprehensible in its existence, but not in its nature.

We Nephologists have much to learn about its nature...

Tuesday, March 17, 2009

The Scale of Cloud Computing

I had an interesting chat with someone at one of the juggernauts in the space today and we were trading stories about Cloud Computing presentations. His comment to me was 'I saw one this morning and it showed 3 servers on it. Do these Cloud guys understand that what they are showing isn't even a test environment in the cloud'.

Exactly why I coined the term Cosmos Computing. The cloud is too small.

We are talking billions of computing components stitched together in some definable but not fully understandable way so that we can use them when we need to. Clouds are but part of and a subset of Cosmos computing.

Stay tuned while I ruminate on this some more.

Quote of the day - Senator Grassley discussing AIG Execs

A prominent U.S. senator has intimated that executives of the troubled insurer American International Group Inc might consider suicide, adopting what he called a Japanese approach to taking responsibility for their actions.

Senator Charles Grassley, the top Republican on the Senate Finance Committee, made his comments on the Cedar Rapids, Iowa, radio station WMT on Monday.

"The first thing that would make me feel a little bit better toward them (is) if they'd follow the Japanese example and come before the American people and take that deep bow and say, I'm sorry, and then either do one of two things: resign or go commit suicide," Grassley said.

"And in the case of the Japanese," he added, "they usually commit suicide before they make any apology."

http://www.reuters.com/article/ousiv/idUSTRE52G3BQ20090317

Friday, March 13, 2009

Cloud Computing is Dead - Long Live Cloud Computing!

I have been on my share of webcasts and seen over 100 presentations about clouds.

You know what they all start with – a definition of what cloud is.

Why?

Is it that confusing? Is it that unclear?

Why is it so confusing and so unclear?

By rehashing and refining a definition, do we hope to somehow impart some clarity to the emerging space?

Cloud computing is unclear. Cloud computing is emerging.

Cloud computing is a paradigm shift and the approach thus far has been to address it like it was not a paradigm or business model shift. The ‘let’s put the Ferrari Red paint on the Apple Cart and call it a Ferrari’ approach.

Why is it a paradigm shift?

Cloud computing is a paradigm shift because there are cosmic forces at work ripping apart, re-forming, and creating new things out of what is and what was every day.

Cloud computing – in my opinion and in its nature - does three major things:

1. Centralizes availability computing resources and components into ‘galaxies’
2. Distributes the galaxies into a collective pool of computing (universe)
3. That universe expands and morphs creating and destroying galaxies

There is the confusion because the definition does not fit the description.

If I am a hardware vendor – I have been in the business of selling hardware. Is hardware the end all be all of computing – no! I need software, networking, and other pieces to make it part of the universe.

If I am a software vendor – I sell the software that provides the ability for me to do something on the hardware, and in conjunction with the network, other software and other pieces in the universe.

If I am a network company – I may build and make hardware AND software, but unless it is connected to other pieces then I am not part of the universe.

So what is cloud computing?

I believe it is not Cloud Computing but ‘Cosmos’ computing.

I really can’t define it in and of itself, but can metaphorically in that it is the equivalent of our Cosmos - universe(s) – much is known, exploration is happening every moment of everyday, and the forces that govern the universe(s) – known and unknown – are at work keeping things in some sort of balance allowing it to work.

I will keep on ruminating on the definition as it goes way above cloud in my opinion...

Thursday, March 12, 2009

Clouds are still...Well...Cloudy

I am catching up on my industry reading this morning and the recurring theme that is (still) enmeshed in the notion of cloud computing is one of cloudiness.

I was on a webcast with Surgient yesterday and bailed after 5 or 6 slides. Why? No useful information was presented. What I mean by that specifically is that the first several slides were all about Surgient and their product positioning. Stuff I could care less about and quite frankly stuff I could get from the 'Products' tab or 'About Us' tab on their website.

Note to webcasters - I need to understand how your product works by you showing me how others have made it work. I do not need to know who the founders are. The obligitory logo slide needs to tell me why these guys use your stuff and how they use your stuff. Answer the question 'So What' in your first 5 minutes or I'm gone.

One of their applications is for setting up 'internal clouds'. What??? Is that the new buzzword to get CFO's to buy new computers?

The 451 Group did a survey in October 2008 and 84% of respondents said they had no plans for internal clouds. If I has to guess why my top guesses are:

1. There is no cloud strategy in place, so no funding for stuff you can't define and understand. See Toxic Assets.

2. CFO's have figured out that a cloud is tech speak for for shiny new computers that will sit underutilized until their leases are up. Again.

3.Why would you build out a cloud for yourself vs. use the ones out there?

One thing that has seemed to gel when Cloud Computing is discussed is that there are generally 3 flavors - SaaS, PaaS, and IaaS. Software, Platform, and Infrastrcuture as a Service respectively. Using these definitions as jumping off points, an internal cloud becomes centralized infrastructure as a service which already exists in some form in the largest companies.

There is much to be done in making the notion of Cloud Computing understandable by CIO's, CFO's and CEO's since they must decide how good it is for their business.

Tuesday, March 10, 2009

I like this stimulus plan better

http://www.microsoftstartupzone.com/Blogs/the_next_big_thing/Lists/Posts/Post.aspx?List=6bab7b08%2D81ca%2D4602%2Dbd97%2D4b7b2c893e88&ID=654

Thomas Friedman of the New York Times today (2/29/2009) wrote “Startup the Risk Takers” where he suggests the US government should stimulate the economy by funding startups, not by bailing out GM and Chrysler.

“You want to spend $20 billion of taxpayer money creating jobs? Fine. Call up the top 20 venture capital firms in America, which are short of cash today because their partners — university endowments and pension funds — are tapped out, and make them this offer: The US Treasury will give you each up to $1 billion to fund the best venture capital ideas that have come your way. If they go bust, we all lose. If any of them turns out to be the next Microsoft or Intel, taxpayers will give you 20 percent of the investors’ upside and keep 80 percent for themselves.”

“GM has become a giant wealth- destruction machine — possibly the biggest in history — and it is time that it and Chrysler were put into bankruptcy so they can truly start over under new management with new labor agreements and new visions. When it comes to helping companies, precious public money should focus on start-ups, not bailouts.”

Friedman’s main point is this; invest in the future (startups), not the past (bailouts). Startups are the best opportunity to create new jobs in high growth industries, and do it fast. According to the SBA small businesses created 70% of all new jobs in 2007, and account for about 50% of all employment in the US economy. Small business is defined as 1-500 employees.

Fred Wilson of Union Square Ventures says “No Thanks”. Fred writes “The venture capital business, thankfully, does not need any more capital. It's got too much money in it, not too little. Just ask the limited partners who have been overfunding the venture capital business for the past 15-20 years what they think. You don't even need to ask them. They are taking money out of the sector because the returns have been weak.”

Government wants to create jobs, VCs want to create wealth – Do VC backed companies create jobs? Yes, but their primary objective is to create wealth. VCs invest about $25 billion a year in over 3,500 companies. These companies grow fast and create jobs. But, VCs primarily fund technology companies with fewer employees and explosive growth potential. Most VCs only fund 2% – 5% of the companies they see, leaving 95% unfunded. These might be the companies that create lots more jobs.

Service based companies create lots of jobs, but don’t get VC funding.- What about all the viable companies that don’t meet the 10X return potential that most VCs demand? Would a VC have funded Wal-Mart, McDonalds, Subway, or ServePro when they were small startups? No way. Service based companies don’t fit the VC model. But, a successful service company might have 5,000 employees, compared to a successful technology company with 500 employees. The government should focus its stimulus money on programs that create jobs. New approaches are needed, but some existing programs are listed below.

Create 50,000 startups for $1B - How about funding for startup incubators similar to Ycombinator and TechStars. These startup mentorship programs are very successful in creating startups and driving innovation but they only cover 30 to 50 companies a year, investing about $20K in each one. Why not start 50,000 companies a year? It would only cost $1B, and could incubate the next Facebook, Google, or Microsoft…and millions of jobs. How many jobs will $1B invested in General Motors create? Government funding for VCs is probably not a good idea, but maybe a small amount of funding for startup incubators would create the jobs that VC backed companies won’t. $1B invested in 50,000 startups would conservatively create 250,000 jobs immediately. If just 2% of those startups became wildly successful they would create 3 million to 5 million jobs.

What can government do? Government should create incentives for investment. It is probably best not to make the investments directly. There are already some good programs and incentives in place that have been forgotten or underfunded for too long. Pouring money into these programs is certain to stimulate investment, inspire innovation, and create jobs. Here are a few examples of programs that stimulate investment from the private sector.

SBIC – Small Business Investment Company A program managed by the SBA, SBICs are privately owned and managed investment funds, that use their own capital plus funds borrowed with an SBA guarantee to make equity and debt investments in qualifying small businesses. The U.S. Small Business Administration does not invest directly into small business through the SBIC Program. This program encourages private investment in startups and small businesses.

SBIR – Small Business Innovation Research Another SBA managed program, the SBIR encourages small business to explore their technological potential and provides the incentive to profit from its commercialization. SBIR targets the entrepreneurial sector because that is where most innovation and innovators thrive. By reserving a specific percentage of federal R&D funds for small business, SBIR protects the small business and enables it to compete on the same level as larger businesses. SBIR funds the critical startup and development stages and it encourages the commercialization of the technology, product, or service, which, in turn, stimulates the U.S. economy.

R&D Tax Credits – There a 20% R&D tax credit for “incremental” R&D spending, over and above what you spent the previous year. There are lots of restrictions on expenses that qualify. Unfortunately startups are limited to a 3% credit because they don’t have lots of prior years spending base. This program could be significantly enhanced by loosening the restrictions, which would in turn stimulate research investment.

Seed Capital Tax Credit – Maine has a 40% Seed Capital Tax Credit for startups, that can go as high as 60% for startups based in high unemployment areas. This is a good example of targeted support for private investment to create startups, jobs, and innovation. More states should adopt Seed Capital investment tax credits.

There are probably lots of other little known programs that support investment in startups. We need to publicize them and invest more in them. Please leave a comment and a link to your favorite program.

Thursday, March 5, 2009

The Story (and Myth) of Metered Power

I was in Dallas, TX this week meeting with a number of companies from Real Estate Brokers to Data Center partners to enterprises. One thing I heard three times was that metered power isn't always metered power - it all has to do with where you put the meter.

At my company we put the meter on the panel that feeds the servers, switches and the gear you house with us. I thought that was the norm and industry standard. Boy was I wrong!

I found out in the course of discussions that some colocation providers put meters on the outside of the building or pod. What's the difference you ask? Big difference!

If you take the base colocation rent that is let's say 100Kw of critical load (draw) and multiply it times .5 and add the 50Kw to the rent to cover cooling and infrastructure, You pay for 150Kw of potential draw and that's what a provider will give you. That's great if the critical draw is 100Kw going to your equipment. Makes sense.

If the meter is placed on the corner of the building or outside the pod, not only are you paying for the critical draw, you are paying for lights, copiers, office space A/C, and everything else NOT related to your gear. Plus 50%. Ouch.

Buyer beware and be sure legal catches this. At $200Kw you could be paying 100-150% of the price per Kw to run signage that's not yours, coffee machines you'll never drink a drop from, and offices that are kept at 64 degrees after hours. Not a good deal, in this or any economic climate.

Tuesday, March 3, 2009

Virtualizationstuff is now cross posted

Just a heads up to my readers - I have cross posted this blog at CRG West and you can take a visit here:

Virtualizationstuff at CRG West

I will remain opinionated, and call it as I see it, so fear not loyal readers...