Our People: Mark's Blog

Episode 65:Zero-Wait Time

I went to check my calendar on my Blackberry the other day and actually got the hour glass! I started to freak out. Had the one appliance that I actually relied on for zero-wait time gotten too complicated? As it turns out, I had loaded some application that was soaking up the processor. So much for that app – I wiped it out!

I need at least one real-time device!

Even with all of the hardware advances, I am blown away by how long I have to wait to do basic things. When I arrive at a hotel and want to sync my email on my laptop, I can easily figure 20-30 minutes to boot, find cables or setup WiFi, configure and pay for Internet, configure and login to VPN, and then sync email. Most of the time, I don’t even bother anymore.

This is still way too hard.  

Talk about what drives value - I will even sit in my office and read my email on my Blackberry because the PC is too slow! I also made the mistake of loading a new version of PowerPoint on my existing laptop (it is only 18-months old). Now I can go get coffee while my files open. After I looked at all of the new features, I went back to my old version because I was so much more productive not waiting 10 seconds for every action!

What scares me is that we are going backwards!

We all like new features. Given the technology advances, I feel we should be able to get new features while not having to actually wait longer in the process. Even my new HDTiVo has a 2- or 3-second longer wait-time to switch between regular and recorded TV. I am not sure the picture quality is worth it.

As far as I am concerned, I have about all the features I can take in my email system and presentation tools – except performance!

Even my new car now takes longer to ‘boot’. There is no reason for this and I think that real-time responsiveness could wind up becoming the biggest new feature around.

Mark…

Episode 64: The Tech Economy – This Time Around

“May we live in interesting times.” I always found this old proverb to be the mantra for the technology industry. It is always interesting, changing and never boring. 

Since the dot-com meltdown, it seems as if every macro-economic woe has a few analysts making dire predictions for technology growth. I will tell you though that, in my opinion, the dot-com bust was a very unique situation and one that will not likely happen again. All industries can be impacted based on macro-economic growth. It’s a simple reality. The times have changed, however, and much of the technology industry is on a new curve.

Back in the now infamous dot-com era, there was a unique dynamic to technology acquisition. Companies (mostly startups and new “Internet” divisions of larger companies) had these amazing business plans that, if they were correct, would require massive amounts of compute, storage and software. Consumption of IT in advance of actual growth became an imperative as growth was supposedly coming so fast that it would be impossible to build capacity once the growth curve hit. The cash in the system made the acquisition part of the plan easy. In short, consumption was inordinately ahead of any real demand. 

As the growth failed to materialize, we were faced with significant overcapacity and the tech market “spiked.” This, of course, was very ugly. If you want to see this spike in detail you can look at the chart of total storage capacity growth over time. It is amazing how consistent this growth has been (in the area of 60% per year) over time. The funny thing is that the growth actually ramped like never before in the late 90s, 2000 and part of 2001 and then only to crash for the two years following as the excess capacity was absorbed.

As people ‘debate’ if we are in a recession or not, I continue to hear all sorts of predictions about the potential impact to the technology industry. While we are tied to macro-economics, for most technology companies this is not even close to what we had to deal with in 2001/2002.

IT capacity today is acquired when it is needed. Software is acquired when there is a definitive ROI (Return on Investment). In short, IT is a business-critical and key productivity asset for companies to maintain competitiveness. New technologies such as virtualization are helping to reduce costs by increasing efficiency. Technology that can help companies be more competitive, reduce costs, and/or reduce business risks is not something that is discretionary. It is even more of an imperative today.

It is funny that when disasters occur, people tend to take action and, in general, there is enough change along with ‘memories’ so that the likelihood of the same disaster happening again is usually pretty low. While we can always learn from the past, it can only serve as a guide to what the future will bring. In this case, we are not dealing with any overcapacity in IT, and it is more critical than ever for companies to seek improved efficiencies, productivity gains and risk reductions.

Mark…

Episode 63: At Number 1 - Information Compliance

 

Time has sure flown by. I can’t believe that it has been two months since my last post. I guess that is bound to happen sometimes. For me, blogging is like exercise. When I get in the groove it is easy but when I stop for a week, I seem to just get into a new groove. OK – on to the topic at hand.

I want to close (finally) on what I believe is the biggest information-centric IT trend for 2008 and that is Information Compliance.  

It is actually quite amazing that this market has taken so long to develop but I think I understand the reason. IT really evolved in very much a bi-polar way. By that, I mean that we had IT come together from two endpoint positions. On one side we grew IT around big corporate systems running OLTP, ERP and such. Typically, these systems are managed in a way that both contemplate and address Information security and protection. While not always perfect, these systems hold information that everyone understand was important, so the systems we designed with Information Compliance in mind.

Over this same period, we started to work and collaborate and do business in whole new ways. Email, chat, the use of the web and other such capabilities moved from interesting tools to business critical systems. The issue, of course, was that most of these applications never contemplated the fact that they would manage such critical information. 

Take a simple example of sending an email to a broker to buy or sell some stock. That is, by most legal definitions, a binding contract. It is also private and confidential. Communications like this are booming simply because it is so efficient and simple. Take an industry like healthcare. Imagine the efficiencies that we could bring to the system if people could communicate more easily with their doctors.

In fact, companies have realized, often in abrupt matter that Information Compliance represents both a risk and an opportunity. On the risk side, all of those emails, IM’s, and file shares represent a potential risk if confidential information is lost or if the information is not managed properly. On the flip side, as we develop more robust ways to protect and secure information, companies and individuals will be able to leverage more of the latest technologies for business critical and confidential needs. 

Information Compliance, by my definition, is an umbrella term that entails that we need to be able to secure protect information both “at rest” and “in motion.” We also must be able to track and audit information and understand the “history.” We must be able to insure both the authenticity and privacy of information. We must be able to control the “rights” that we want to give to individuals relative to a pice of information. We must have business rules and retention policies that insure that we have retained all of the necessary information. Finally, we must be able to “discover” information that may exist are hundreds of applications, many thousands of people, and many billions of objects pertaining to a particular topics.

As a whole, this is what information compliance is all about. 

For many companies, putting more information compliance capabilities in place will be simply about reducing risk. Over time, this capability will let us do more with our information. We will be able to share it in more secure ways and interact with more efficiency and less bureaucracy.

Mark….

Episode 62: Number 2 - Green Computing

Going Green

I was touring a datacenter at a customer recently and walked into a huge room that was maybe 1/8^th full. The IT manager noted that this datacenter was now “at capacity” and no more systems could be added. He commented that the industry has been so good at increasing the power density that most of the existing data centers now run out of power when the are only 25% “full.” Today, new data centers are being built with 3 to 4 times the power and cooling capacity of existing facilities. They are also being built in locations that can be better served with massive amounts of power (e.g. near big power plants) because it is easier to move bits (data) than atoms (power).

To put it bluntly, green computing is not a fad; it is a pervasive trend based on simple economics. The economic reality is simple; the cost of power is now a major driver of datacenter operational costs. 

Storage systems are a big power consumers and, unfortunately, they have gotten exponentially worse as we have increased the speed of disk drives. Unfortunately, when we increase the rotational speed of a disk, power consumption increases exponentially. Basically, doubling disk performance (in terms of rotational speed) effectively quadruples power consumption – not a good thing.

Green Technology

While we will always work to make systems more efficient in terms of power consumption, I believe the big gains will come from several specific technologies. 

The most obvious of these is virtualization. Virtualization has enabled dramatic server/application consolidation. Going forward, with even better tools to enable dynamic load management, we will truly be able to dynamically power off and on server capacity. For the few people that haven’t networked and separated severs and storage, your costs will continue to rise as virtualization simply will not be effective.

Information Storage is a completely different animal. Information (capacity) does not grow and shrink with load and it is difficult if not impossible to completely “turn off” access to information. Much can be done however by using new techniques and technologies that can dramatically reduce the power load and still insure access performance and availability. 

One of the new trends will be the use of high-reliability flash to create a new high-performance/low power storage tier – "Tier 0" as it is often called. This is not the flash in your iPod (which has a life of maybe 10,000 writes); it has much higher cycle life, making it practical for enterprise storage applications.  Flash of this type is not cheap but it can have a great ROI for frequently accessed information, especially when the power costs are considered.

Another trend will be the increased use of bulk storage tiers that leverage high capacity medium-performance drives. In many cases the drives will be spun down when not in use. Access will be slower but it will still only take “seconds” to get the information. In many cases, such as with archives, this will be more than sufficient.

I am often asked if we will now store more data “offline” on tapes, DVD’s,etc. My simple answer is no. The cost of manually managing physical media and long access times make these repositories impractical. 

So, while technologies like virtualization, flash, and even information lifecycle management (storage tiering) were not invented specifically for energy conservation, they are clearly being adapted to address this need. We (the IT industry) are fortunate because I believe this will be one of the key areas of IT investment for 2008.

Mark….

 

 

Episode 61: Rich Media Goes Digital―Really!

I just returned from my first trip to the XGames. I went at the invitation of my friends at ESPN. In addition to being one of the best sporting events I have ever attended, I was able to go behind the scenes to see the production crew and systems in action. I will tell you – while we all know that ESPN is the clearly the place for sports – what you may not know is how much they are also a technology leader and how much those capabilities provide for a differentiated customer experience.

I won’t go into any detail about all the innovative things going on (the ESPN folks may want to keep those to themselves) but I just want to highlight one key thing. In the entire tour of the event and the facilities I did not find or see a single tape, cassette, or piece of removable media. Not only was everything digital, it was all effectively created, produced, and distributed within a single digital information infrastructure.

If you have been in this industry for any length of time, you have been through at least one “false start” in terms of Digital Rich Media. Back in 1990, I remember hearing that by 1992 we were all going to “Digital Video on Demand” in the home. Of course, it never happened. The marketing hype continued, and consequently, a kind of “Chicken Little” philosophy has developed around the Digital Rich Media market.

Well, my prediction is that we all better be prepared for the sky to fall because the all-digital rich media market is truly upon us.

Yes, I know that much of what we see on TV today is “digital.” Some TV and movies are being shot and produced digitally as well. Let’s not forget to mention the amount of “digital” content on sites like YouTube. So, let me explain what I mean when I say thatrich media goes digital.”

Up to this point, the development and production of rich media has been an amalgamation of analog and digital processes with many of the digital processes using custom HW to provide a given set of functionality. Transformations, from creation to editing, to review to production and archival were largely what I call mechanical processes. A person would take a tape (or maybe a disk) from one place to the next or one machine to the next.

The future is the true convergence of IT and Digital Rich Media in a way that the most cost effective (and existing) IT systems infrastructures can be used for the end-to-end production and distribution of Rich Media Content. It is not just about making Digital Movies or TV broadcasts; it is about leveraging the power and cost savings potential of standard IT infrastructure to transform the market. This is possible, in large part, because the rapid capacity and performance improvements in IT systems are not reaching the threshold of capability that will yield the necessary performance and reliability for Rich Media.

The merging of commercial IT with Rich Media will bring about a whole new set of capabilities for consumers and economic potential for producers. First and foremost, the basic economics will be transformed with the use of IT systems. Additionally, process and procedures will become more streamlined and more automated. Distribution will become more efficient and flexible. Personalization (and therefore monetization) will improve exponentially.

This will all happen with video rich media with the security and rights management equal or greater than what is available with the present media-based (e.g. DVD) methodologies.

You can see this starting to happen today in so many places. Our kids are not watching TV on the TV anymore (they watch it on their laptop or phone). Video rentals are delivered to our iPhone or Tivo. We are watching more and more niche content directly on the web. This goes on and on.

So, I believe that the next revolution for Rich Media will not involve any media at all! From creation to editing, to distribution and viewing, our content will never touch a DVD, removable disk, or tape.

Mark…

Episode 60: Number 4 - Unified Information Management

Coming in at number four on my list for 2008 is “Unified Information Management.” In context with many other elements of IT and communications today, we have developed a myriad of strategies for managing our information. Many of these have evolved with individual deployment of applications. At the time, it made sense to deploy separate information management strategies because the information used by applications was largely silo’d, and the data management costs were allocated to individual applications. 

The communications market is going through a similar transformation. At one time companies viewed voice, video, email, and data communications as completely separate elements with little synergy or leverage to be gained (or possible) from unification. Today, however, these communication elements are coming together in a way that can drive both increased productivity as well as reduced cost.

From the standpoint of information, we can see a similar dynamic happening. Applications are no longer silo’d, they are connected and “mashed up” to support more advanced capabilities.  The data itself is now shared across applications. Additionally, there are more and more regulations governing the security, protection, and retention of information.  Welcome to Web 2.0. 

So, what looked like a good idea at the time (application-specific information management) is now actually hurting a company’s ability and effectiveness for managing information properly.

Unified information management is simple. Companies will build a single intelligent infrastructure platform to protect, secure, and manage information. These information infrastructures will support all applications across the IT environment. A similar analogy holds true for even our personal data. While we may have different requirements for our application data (music, photos, records) I would still prefer a single infrastructure for all of my data. This infrastructure will have all of the necessary capabilities like backup and ILM; these capabilities will be deployed based on policies that insure the proper levels of protection and security.

Mark...

Episode 59: Top 5 Picks for 2008

As we approach the end of the year, it is a good time to both reflect on 2007 and think about what is ahead for next year. As usual, there is a good deal of debate about IT spending trends. What gets less debate, but is likely more important however, is how the industry is positioned to deliver technology innovation to meet new business needs and drive value.

The simple fact is that IT spending priorities will align with the segments where technology innovation can drive the largest return on investment. Overall, I believe that building heterogeneous information infrastructures will be a key priority for organizations in the coming year. So, with that in mind, my next blog entries will discuss my top 5 picks for the hot trends relative to information infrastructures and IT for 2008.

Number 5 - SaaS for Information Services

Software as a Service (SaaS) has become a key technology for deploying business applications. SaaS applications can provide excellent value by lowering up front costs, and speeding deployment. I believe this trend will continue to evolve with more services becoming available that expand the use beyond just specific applications. One example is data backup. Our Mozy SaaS backup solution is leading the way to help consumers backup their personal data with a simple, complete, and very cost effective transformational solution. Look for these infrastructure SaaS capabilities to have more and more applicability to businesses and corporations. 

I also expect that we will see the emergence of what I call generation 3 SaaS platforms. In the first generation of SaaS, many applications were simply repackaged as hosted applications. They offered a more flexible business model but were not cost-optimized. The second generation of SaaS has been driven by the emergence of dedicated SaaS platforms. This step greatly increased the effectiveness of the offerings because the products were developed specifically for multi-tenant (shared) environments. These applications tend to be all-or-nothing, however, and do not allow a great deal of flexibility.

Generation 3 SaaS applications will incorporate multi-tenancy and also incorporate the ability to run more seamlessly across onsite and offsite datacenters to more effectively serve enterprise environments. A combination of virtualization as well as classic multi-tenant platforms will be used to deploy customized applications across a shared platform. Within a common “information platform” businesses will be able to deploy both on-premise and remote applications.

Exciting stuff...

Stay tuned for the countdown to number one.

Mark…

 

Episode 58: Rules Rules Rules

There are now more than 20,000 laws world-wide that regulate how companies must protect, retain, and secure information. These laws are often complicated, and sometimes even contradictory. These rules often must be applied to all data types regardless of the application. In many cases, information cannot even leave the country in which it was created. Additionally, companies themselves are seeking to put in place policies for data retention and disposition in an effort reduce their own risks. 

It sounds complicated and it is. But, it gets worse. Most companies have hundreds of home grown applications, file shares, and records repositories; not to mention Email and ERP applications.  If that isn’t enough, most large companies today operate globally so they must be compliant with many unique local regulations.

OK.  If a typical multinational company operates in 50 countries and had to apply just 50 regulations and 10 corporate policies across 100 data repositories (applications) the simple task would be overwhelming. They would have to process upwards of 300,000 unique rules! 

So how do companies deal with this complexity? The fact is that I believe that most don’t. This is an almost impossible problem in most current IT environments. To solve the problem I believe we will fundamentally need to change the way we look at information. The secret is decoupling information from individual applications and creating unified and federated content repositories where rules and policies can be applied regardless of the application. The simple math is that, by unifying the repositories in the above example, the number of total rules drops by a factor of 100.

This does not mean that all of the data moves to one big archive or repository. On the contrary, much of the content created must to remain in place. Federation, simply means that there is a uniform way to set data taxonomies and data policies so that requirements can be applied in a simple and uniform way. 

Look at it this way; imagine that every state in the US or country in the word had completely unique road signs – so different in fact, that is was impossible to drive from one state to the next without learning a whole new set of rules. But it get worse, imagine every car you drove had a different measure for it speed. This is what we face today. With common taxonomies (in this case road signs and gauges) universal and local rules can be easily applied and there is a consistent understanding of the data. Companies today are faced with a similar information challenge. The key is to first build common taxonomies and then to build policy.

You will see that our efforts relative to information compliance focus on our core strategy of first building a common information taxonomy framework. This is the secret to taking this incredibly complex problem and cutting it down to size. With an information compliance strategy and an cross-application information infrastructure I believe customers can what is today a liability into a competitive asset.

Mark...

 

Episode 57: Information Platforms

There was an interesting article in the WSJ the other day about “platforms” and how companies can succeed or fail based on their strategy and actions. While it is never as simple as just having a few rules, the basic arguments are sound. Overall, platforms need a high degree of openness around them. They also require ways for others to both compete and add value.

In the past, many “platforms” were based on technology like DVD’s, Windows or Mainframes. More and more, we are seeing the rise of what I would call web “information platforms.” Most social networking sites are simply information platforms. Their value as a platform is not in the technology itself; they provide value because they have a collection of information.  Platforms naturally evolve here simply because there is greater information leverage by being able to operate across the most complete set of information. No matter if it is searching the web, connecting with others, or buying and selling stuff, there is value in the depth of information found in platforms that have scale.

What is a bit ironic is that many if not most companies actually are behind in building information platforms for their own information. It is not because they don’t want to or because they would fail to add value. On the contrary, the same leverage that we see from information platforms on the Web can be applied to leverage corporate information.

I believe the primary reason why we are where we are is simply that most corporate IT systems evolved in what I would call the “application era.” The focus at the time was on building a given application and, principally, data was considered to belong to one, and only one, application. Consequently applications evolved to use silos of information and we have been working ever since to integrate applications to try to improve business processes.

What I believe is required, however, is to shift focus towards creating information platforms for business data. Unfortunately, there are also more complications to creating business information platforms. Unlike much of what we use on the web, corporate information must remain secure and often private. Information use often will need to be tracked and meet various regulatory requirements. Notwithstanding, I believe that there is great value in companies building information platforms.

Building corporate information platforms is not going to happen overnight and it is not just about putting information in a single place. To be effective, companies will need to adopt and information-centric view towards their overall IT architecture. This means that information will need to have security and protection built in at the data level. It also means that more metadata will need to be maintained and tracked.

Clearly at EMC, our primary strategy is to help our customers build information infrastructures which will then enable the building of information platforms. Imagine the leverage a company can get by having all of its information available to any application for any valid use. I believe the competitive leverage is enormous!

Mark…

 

 

Episode 56: Innovation Conference 2007

It has been difficult to get enough time to write lately. I am that middle of that job change time where I seem to be doing two full time jobs - finishing my commitments as CDO while fully engaging with the new team. What is really hard is that there is so much going on to talk about. 

Today, EMC is holding it first “Innovation Conference.” It started as an idea to bring innovations and innovators from across the organization together for a conference specifically dedicated to innovation.

We had Don Tapscott (the Author of Wikinomics) speak at the conference this evening. Don is a great speaker and gave the team a compelling set of arguments as to how new forms of collaboration will change how companies operate. 

To me, the one thing I have learned as a strategist (and sometimes inventor) is that that ideas come from everywhere. If you relegate you idea generation to any small group of people you will not get the best results. Mass collaboration becomes the critical element to generating the best possible ideas and solutions.

In test past, we were effectively “technology limited” in our ability to collaborate on a wide scale. Email was just not going to get us there. Today, however, we have the tools to truly collaborate on a global scale. 

The other key element, to me, of innovation is the ability of an individual or company to turn ideas into real products – implementation if you will. Innovation is really invention + implementation. Here, companies must provide both the support and establish a culture that fosters innovation.

More on the conference as it continues…