E2K: The Additional Constraint

In 1965 Intel co-founder Gordon E. Moore first observed that the number of transistors that can be inexpensively placed on an integrated circuit will increase exponentially, doubling approximately every two years. That time he himself failed to estimate the longevity of this powerful statement and felt uncertain of the any projection beyond 10 years. Half a century later, the computing power is still growing at the same rate and we find ourselves at the same cross road when it comes to predicting the longevity of his observation.

This increase in computing power fueled many dreams. Small chips computing at Giga Hertz scales made devices go Nano. This combined with tremendous increase in storage availability and high internet bandwidth opened up opportunities just waiting to be dreamt about. Cost of Hardware went relatively down and many luxurious objects suddenly became very affordable. Google said that its mailbox size availability will expand every second and there would be no need to delete mails! In reply Yahoo came out with an unlimited mailbox for all. Communication devices went dirt cheap. VoIP became so affordable.

The rate at which world is changing seems not even exponential but beyond. Soon we can hope for surface computing devices where any surface will have amazing computing power. The colour of the office walls will swing daily, the restaurant table will act as personal desktop through which one can order food or surf net or download pictures or transfer documents. There will be a MindTree's space in second life with a 3D office open for anyone to visit and check for available facilities. CXOs laptop will be of the size of a pen projecting 3D images in the air when started. What luxury!

Dreams seldom have limitations, unlike the real world. And suddenly we realize that in mid of this luxurious imagination, we have kind of forgotten the additional constraint. As computing power increases, we jump upon to use it. It results in above-exponential rise in number of transactions made. Transactions consume energy. While we are building robots to run faster, what we are essentially forgetting is that for running faster they would need more food. Food here is the electricity these devices consume. While we dream of those luxurious computing power and ever-available disk spaces, we fail to notice the way these chips and disks are guzzling our fossil fuels.

Efficient markets are said to be the best mirrors. Check any commodity exchange. Oil prices ended the year near $96 a barrel, or 57 percent higher than where they began. This is considered to be a record-breaking year for energy futures. This is the biggest percentage gain since 1999 when prices more than doubled to $25.60 a barrel. IT industry today emits as much CO2 as does Airline Industry. But the reasons for this rise in Energy prices may be much beyond IT. The point to be noted here is that every increase in Energy costs adds indirectly to the total cost of ownership of computing power. This comes up as an indirect cost and corporate houses may get caught unaware of the implications of this indirect parameter which may increase its significance by 50-60% per annum. Organizations may easily tend to over-provision and underutilize their IT infrastructure due to easy availability of cost-effective processing power. This will be akin to owning white elephants who would keep on guzzling high energy resources even when they are just a luxury. This luxury will pinch pockets as energy prices soar above the roof.

In many ways this constraint has a disruptive character. As world wakes up it will look forward to dynamic optimization of resources to lower down its total cost of ownership. Many parameters will play part in these optimization algorithms. Focus will be on maximizing usage of every bit of energy produced. Processing power will be a choice and not a dream and may be amply reduced. Data redundancy especially at large banks will be a pain. Governments will be much harsher on energy usage. Thin clients, distributed computing, server virtualization, carbon footprints, energy efficiency certifications, etc will be the buzzwords.

Thomas L. Friedman who found the world as flat as Infosys’ huge Video conferencing screen calls E2K as the next Y2K. Though E2K does not have that strict Y2K kind of deadline, still the impact will be huge. It will be imperative for the world to support it. Few big MNCs like IBM and EDS have an inherent advantage due to their solid presence in server business or IT infrastructure business. But in the end Indians will be writing most of the applications!