The Grid
The Grid Computing category of Ross Mayfield's Weblog

Subscribe to "The Grid" in Radio UserLand.

Click to see the XML version of this web page.

Click here to send an email to the editor of this weblog.


Thursday, May 22, 2003

Value in Aggregate

Clay Shirky's latest newsletter, on Grid Supercomputing: The Next Push, devalues computation relative to communication and questions the value proposition of grid computing.  He makes some great point, but I think he is right about relative value, albeit the irony that grid computing is enabled by communication. 

When I think of grids, I think of bell curves.  MIPS, Mbs & Mbps have become fungible datacommodities.  Grid computing provides means of aggregation.  When aggregated you pool risk.

Your average company utilizes datacommodities in a bell curve that maps to the duration of a day, with a peak utilization a little before noon.  This bell curve follows the sun in a daily cycle.  A company in California has a different demand curve than one in NY, creating an potential overlap in preferences.  Grid computing allows those preferences and risks to be pooled, flattening the curve in aggregate.  The key to making it work is the transaction costs of shifting granular supply from one place to another, billing for it and managing its performance.  Virtualization technologies in grid computing and data centers are advancing to make these otherwise smart services dumb.

Californians became intimately acquainted with the danger of bursts at the peak, falsified or not, in the electricity market.  Here is an example of how pricing is differentiated for load aggregation according to base, peak, off-peak and burst:

Sure processing is relatively inexpensive and has Moore's Law on its side.  Grid computing isnt for individuals.  When costs are added up for a large organization, hosting farm or major project they are significant.  Pooling costs and risks in the aggregate has its benefits.

IBM's marketing whizes have it right in promoting the on-demand potential of grids.  Nothing panics decision makers more than the risk of not having enough capacity available at a key unforseen moment.  They will pay a premium for capacity on demand or as an insurance policy.

8:04:51 AM    comment []

Monday, May 19, 2003

The Matter of IT

As the technology industry emerges from the bottom, it criticism has been lowered to the nature of being.  Information Technology has always been a source of competitive advantage.  At the bottom of the business cycle, advantage is basic survival, risk taken only on low hanging fruit.  But now, as the leaky pipes have stopped dripping, some wish to relegate the job of IT to that of a plumber.

Friday's article in the NY Times by Steve Lohr asks the question, "Has Technology Lost its Special Status?"  Technology as an industry is 10 percent of the economy and 60 percent of business spending.  According to John Gantz of IDC, technology spending has increased 2-3 times the rate of economic growth since the 1960s.  The question is if the industry will return to its historical growth rates, the driver of valuation multiples.  Lohr reports:

That assumption about technology's special role is questioned in a provocative article this month in The Harvard Business Review, titled "IT Doesn't Matter." The article asserts that information technology, or I.T. for short, is inevitably headed in the same direction as the railroads, the telegraph, electricity and the internal combustion engine — becoming, in economic terms, just ordinary factors of production, or "commodity inputs."

"From a strategic standpoint, they became invisible; they no longer mattered," Nicholas G. Carr, editor at large of The Harvard Business Review, wrote in the article. "That is exactly what is happening to information technology today."

John Hagel and John Seely Brown believe this is an important article because it very effectively captures the backlash sweeping through executive suites against IT spending...But Carr’s article is also dangerous because it endorses the growing view that IT offers only limited potential for strategic differentiation....and are preparing a rebuttal in the July issue of HBR:

  • Extracting business value from IT requires innovations in business practices. In many respects, we believe Carr attacks a red herring – few people would argue that IT alone provides any significant business value or strategic advantage.
  • The economic impact from IT comes from incremental innovations, rather than "big bang" initiatives. A process of rapid incrementalism enhances learning potential and creates opportunities for further innovations.
  • The strategic impact of IT investment comes from the cumulative effect of sustained initiatives to innovate business practices in the near-term. The strategic differentiation emerges over time, based less on any one specific innovation in business practice and much more on the capability to continuously innovate around the evolving capabilities of IT.

Some tech executives have countered in the NY Times article:

"I.T. is the vehicle by which you turn ideas and content into intellectual property products," Mr. Craig Barrett (of Intel) said. "As a nation and as a company, you either upgrade your I.T. infrastructure or you won't be competitive."

Samuel J. Palmisano, chief executive of I.B.M., made the case for his industry's growing at twice the rate of the economy when he spoke to analysts on Wednesday. "The industry is fundamentally a growth industry because it underpins productivity," he said.

Kevin Werbach speaks of new models in the Post-PC era in his Supernova report:

...The point of the article is not that tech is dying, or that innovation is drying up. It's that enterprise technology is moving into a new phase. Bigger, faster, and more feature-laden are no longer selling points in the same way. Smarter, simpler, more efficient, and more flexible are the new criteria. It's much harder to make powerful system simple than to make them complex....

Zack Lynch points how IT drives growth in other sectors:

For instance, although not a punctuated leap in competitive advantage, social software has the potential to accrue significant value for companies that leverage its potential to accelerate innovation. In some industries, two product cycles can be the difference between corporate life and death.

For example, decreasing innovation cycle times in the pharmaceutical industry by 10% could slash years off product research, development and approval processes.  When translated into revenue and market capitalization impacts, intelligent adoption of social software could significantly disrupt the balance of power in this multi-billion dollar industry.  Who says IT competitive advantage is dead?

What is unique about IT compared to other revolutions is how it extends the capabilities of people and groups.  IT fuels competitive advantage by enhancing productivity.  Erik Brynjolfsson of MIT has demonstrated that productivity gains occur not through IT in and of itself, but when it is introduced with new business practice and process.

This is actually a contrarian point for many.  Some private equity investors shy away from technologies that require change of behavior, for example, because it adds risk that users will resist change.  But in fact, that's the real promise of IT, to extend our capabilities in new ways.  Changing behavior is good in a changing world.

Brynjofsson's studies have been in process-intensive areas of organization.  Areas where economies of scale can be easily realized.  When business process is defined, it is almost immeadiately outdated because environmental conditions change presenting new sets of information. This underscores the need for business practice that realizes economies of scope (agility), but also the limits of process itself.

What remains is knowledge work.  Most jobs in the service sector spend the majority of their time undertaking unstructured tasks in social context.   And this time is where innovation occurs (Palmisano points to IP creation, but that's just one set of montetizable outcomes). 

IT will continue to drive competitive advantage for business because an incremental enhancement of how groups work still yields exponental benefits.

11:38:47 PM    comment []

Thursday, March 06, 2003

World of Ends

Dr. Weinberger and I decided to sum up a whole bunch of stuff in one big site: World of Ends: What the Internet Is and How to stop Mistaking It for Something Else. Dr. W. explains more here.[The Doc Searls Weblog]

The Nutshell

1. The Internet isn't complicated
2. The Internet isn't a thing. It's an agreement.
3. The Internet is stupid.
4. Adding value to the Internet lowers its value.
5. All the Internet's value grows on its edges.
6. Money moves to the suburbs.
7. The end of the world? Nah, the world of ends.
8. The Internet’s three virtues:
a. No one owns it
b. Everyone can use it
c. Anyone can improve it
9. If the Internet is so simple, why have so many been so boneheaded about it?
10. Some mistakes we can stop making already

It all begins with Simplicity, turns out bandwidth is a commodity, and let's be stupid and not screw it up.

10:41:48 PM    comment []

Saturday, February 15, 2003

Google buys Pyra

Google Buys Pyra: Blogging Goes Big-Time
posted by Dan Gillmor 07:41 PM
permanent link to this item

NOTE: This is a slightly edited version of a special column running in tomorrow's San Jose Mercury News. We're posting it early to get the story out.

Weblogs are going Googling.

Google, which runs the Web's premier search site, has purchased Pyra Labs, a San Francisco company that created some of the earliest technology for writing weblogs, the increasingly popular personal and opinion journals.

The buyout is a huge boost to an enormously diverse genre of online publishing that has begun to change the equations of online news and information. Weblogs are frequently updated, with items appearing in reverse chronological order (the most recent postings appear first). Typically they include links to other pages on the Internet, and the topics range from technology to politics to just about anything you can name. Many weblogs invite feedback through discussion postings, and weblogs often point to other weblogs in an ecosystem of news, opinions and ideas.

"I couldn't be more excited about this," said Evan Williams, founder of Pyra, a company that has had its share of struggles. He wouldn't discuss terms of the deal, which he said was signed on Thursday, when we spoke Saturday. But he did say it gives Pyra the "resources to build on the vision I've been working on for years."

Part of that vision, shared by other blogging pioneers, has been to help democratize the creation and flow of news in a world where giant companies control so much of what most people see, hear and read. Weblogs are also becoming a valuable communication tool for groups of people, and have begun to infiltrate the corporate, university and government spheres.

Just three and a half years old, Pyra's Blogger software has 1.1 million registered users, Williams said. He estimated that about 200,000 of them are actively running weblogs. Pyra charges for some higher-capability services not available in the base configuration, but most of its registered users don't pay.

Google is known best for its search capabilities, but the Pyra buyout isn't the company's first foray into creating or buying Internet content. Two years ago Google bought, a company that had collected and continued to update Usenet "newsgroups," Internet discussion forums. More recently, it created Google News, a site that gauges the collective thoughts of more than 4,000 news sites on the Net.

But now Google will surge to the forefront of what David Krane, the company's director of corporate communications, called "a global self-publishing phenomenon that connects Internet users with dynamic, diverse points of view while also enabling comment and participation."

"We're thrilled about the many synergies and future opportunities between our two companies," he said in a statement on Saturday. He didn't elaborate further on what those synergies and opportunities might be, but said more details would emerge soon. Users of the Blogger software and hosting service won't see any immediate changes, he added.

For Williams and his five co-workers, now Google employees, the immediate impact will be to put their blog-hosting service, called Blog*Spot, on the vast network of server computers Google operates. This will make the service more reliable and robust.

How Google manages the Blogger software and Pyra's hosting service may present some tricky issues. The search side of Google indexes weblogs from all of the major blogging platforms, including Movable Type and Userland Radio. Any hint of proprietary favoritism would meet harsh criticism.

Blogging was moving mainstream even before this buyout. Several weblogs draw a large readership, and bloggers demonstrated their collective power to keep an issue alive even when the traditional media miss the story, as former U.S. Senate Majority Leader Trent Lott discovered to his dismay late last year.

Major technology companies are seeing the potential. Tripod, the consumer web-publishing unit of Terra Lycos, recently introduced a "Blog Builder" tool. America Online is expected to be on the verge of doing something similar, and no one will be surprised if Yahoo and Microsoft do the same. Are more buyouts of blog toolmakers in the offing?

Developers of blogging software have been finding user-friendly ways to help readers of weblogs and other information find and collect material from a variety of sites. It's in this arena that the Google-Pyra deal may have the most implications.

More than most Web companies, Google has grasped the distributed nature of the online world, and has seen that the real power of cyberspace is in what we create collectively. We are beginning to see that power brought to bear.


9:29:04 PM    comment []

Thursday, January 23, 2003

Distruptive Technologies

10 Emerging Technologies That Will Change the World. dsg writes in with a link to this Technology Review article about ten emerging technologies they believe will change the world. Some interesting predictions in the bunch. Mixed in with obvious predictions like "grid computing" are slightly more obscure things like "nano solar cells". These types of articles are always popular in January, and while no one ever seems to look back and see how accurate they are, they're still fun to read just to get your mind thinking about the possibilities. [Techdirt]

I like these lists for the same reason.  Here are the 10:

  1. Wireless Sensor Networks
  2. Injectable Tissue Engineering
  3. Nano Solar Cells
  4. Mechatronics
  5. Grid Computing
  6. Molecular Imaging
  7. Nanoimprint Lithography
  8. Software Assurance
  9. Glycomics
  10. Quantum Cryptography

Creating blood vessels is a huge challenge (Image: SUSUMU NISHINAGA/SPL)

In other news, we are printing cells...

Three-dimensional tubes of living tissue have been printed using modified desktop printers filled with suspensions of cells instead of ink. The work is a first step towards printing complex tissues or even entire organs.

"This could have the same kind of impact that Gutenberg's press did," claims tissue engineer Vladimir Mironov of the Medical University of South Carolina.



11:48:06 AM    comment []

Monday, December 09, 2002

Datacommodity Service

IBM rolls out on-demand computing service aimed at supercomputers.  [InfoWorld: Top News]

Selling processing as a utility service (one of the Datacommodities) from a grid computing architecture is finally commercialized.  IBM's first customer, Petroleum Geo-Services, expects to save $1.5 Billion for a three-month-long seismic imaging project based in the Gulf of Mexico.

"Customers in some sectors want access to large-scale computing power in short bursts," said Dave Turek, vice president of IBM Linux clusters and grid solutions. "We think this supercomputing offering can change how business is done."

Aggregating peaks and valleys into a plain is a great business.  Especially when customers will pay a premium for serving bursts while not having to incur capital expenses at the peak.  This will be one hell of a spot market one day.

11:19:37 AM    comment []

QoS vs. Open Standards

Cory on Rod Smith:

Open standards and quality of service: pick one. IBM's Rod Smith is speaking at the Supernova conference. In his intro, he cites a lot of customer demand for both open standards and quality-of-service guarantees. Aren't these antithetical? If I'm running open standards, then the software at my end of the network can be set to abide by or ignore any signals send by the software at your end (as opposed to a proprietary system where both ends are welded-shut-boxes that always and deterministically do whatever the software author thought was best). That means that even though your software requests a priority level of x and a guaranteed pipe of y, you have no way of knowing whether my software is actually delivering x and y. All you can send me is a suggestion -- not a guarantee.

Not mutually exclusive.  With two clients running open systems and each buying their own service level agreements, the clients could act in a manner similar to cognitive radios with open spectrum, but that's theoretical.  But what Rod was talking about was not end-to-end, but enterprise-to-enterprise (which then extends it intra-enterprise to the end, mostly machines for grid computing).  This is accomplished via:

  1. VPNs with QoS guarantees, perhaps leveraging MPLS at least in the core of the service provider network
  2. Companies buying near wholesale IP Transit agreements with traffic specific SLAs.  Each leverages multi-homing that takes into account the cost of Transit and SLAs for each provider.  Or is there an open standard Im missing?

What's not open about this all yet is standardization of tiers of QoS from Layer 3 and up.  But layer 2 and down is defined by ANSI (a DS-3 is a DS-3).  Or is there something Im missing.

11:18:53 PM    comment []

Rod Smith, IBM

Fascinating to me, but much of IBM's new commodity management model, wasn't understood enough to be explored by the audience.  So the focus was from web services to the consumer in the value chain.  The missing piece was data commodities to web services.

On Demand

  • Responsive in real-time
  • Variable cost structures
  • Focused on differentiating
  • Resilient, Global

Drivers for Next Gen Biz Apps: inter-enterprise, integration & QoS.  Coordinating decentralization


  • Integrated
  • Built on open standards - what's new is standard process is quicker
  • Virtualized
  • Autonomic

Question of QoS tiers for commoditization.  This is not an issue, IMHO -- there are over 150 grades of West Texas Intermediate Crude WTI -- the largest oil commodity contract.  I doubt anyone in the room gets commoditization, but they will.

Great questions in the summary slide:

  • What if integration costs went to zero?
  • What if IT disolves into the fabric of a company?
  • What happens when integration decisions happen at <web service> deployment time or connection time... or business contract time?

The real opportunity is accelerating the Innovation-Integration cycle. 

Floor questions:

  • Tom from UCB asks for examples of web services that delivers business value today.  WebBeacons took an internal app and ASPed it to eliminate 15 people who manually proceessed their trucking procurement 3 days to 4 hour turn around time, reduced cost by $1m.  e2open, UDDI and SOAP, now up to 600 companies integrated. 
  • Dave Winer asks about Patents again.  IBM has not patented what's in SOAP. 
  • Follow Marc's question on patents,  IBM asks startups they work with to adhere to open standards (which costs them considerably) -- reasonable answers.
  • Isen: IBM had the world by the short hairs, but is now relatively decentralized.  How do we help Microsoft change like IBM did?
    • language change
    • connected with customers to get their view of the issues

5:41:46 PM    comment []

Sunday, December 08, 2002

Waves and Things

Mitch Ratcliffe puts the Next Big Thing list in perspective...

While I can appreciate the focus that a "big thing" lends to folks with a business degree, the broader and more flexible approach to decisions about resources afforded by a historical perspective tell me that it is the Next Long Move, a trend that pulls people and organizations along new paths, that provides the richest opportunities. Trends are based on collections of things, not any one thing... A focal point-like "big thing" makes you subtract everything from the picture that doesn't fit. [RatcliffeBlog]

Schumpeter's Business Cycles (1939) proposed a three-cycle model of economic fluctuations or waves. Squeezing a fourth cycle between his second and third, we get...

  1. Kitchin inventory cycle (3-5 years)
  2. Juglar investment cycle (7-11 years)
  3. Kuznets infrastructural investment cycle (15-25 years)
  4. Kondratieff long cycle (45-60 years)

For Schumpeter, three Kitchins make up one Juglar and six Juglars make up one Kondratieff. Fitting in the Kuznets, we presumably have two or three Juglars to one Kuznets and three Kuznets to one Kondratieff.

Right now we are at the beginning of the next Juglar investment cycle.  I believe what Mitch calls a Long Move is a Kuznet wave, a major adoption of new infrastructure that impacts all facets of humanity.  We are in the middle of a Kondratieff wave, that of IT. 

The next Kondratieff wave is what Zack Lynch is blogging about, the Neurotechnology wave.  Zack speaks in waves, so if you want to learn more about the concept and history of waves, follow his posts.

I agree with Mitch that Things are not enough to base an cash or time investment decision upon, larger trends, or waves, need to be taken into account.  However, I like lists of Things to be sure Im not missing any I need to invest time to understand.  And when you find something that interests and makes sense for you, understand the wider context -- and go for the Juglar Jugular.

5:12:56 PM    comment []

Next Big Things

John Patrick's list of five candidates for the Next Big Thing. [Werblog]

  Autonomic computing
  bullet Blogging
  bullet Grid Computing
  bullet Web Services
  bullet WiFi

John is about right, IMHO.  I have been trying to think in non-Next Big Thing terms, but would add: Social Software, Personal Systems, and more to come.  The good thing is we will be talking about 3 out of 5 at Supernova, with the other two in the hallways.

9:59:10 AM    comment []

Click here to visit the Radio UserLand website. © Copyright 2003 Ross Mayfield.
Last update: 5/22/2003; 8:05:39 AM.
This theme is based on the SoundWaves (blue) Manila theme.
May 2003
Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30 31
Mar   Jun