Keys To The Digital Future.

The information age is now in full swing. The ones and zeros that computers create, process, store, and retrieve are the core fabric of digital data. To be sure, we can represent numbers, letters, text, speech, video, pictures, graphs, 3-D images, and the state of being on or off, true or false, in or out, and up or down just with ones and zeros. Is the digital storage explosion that we are experiencing defined at its simplest level to be the business of managing ones and zeroes? In reality, it is much more than that. Much of the world's business now runs on Internet time. This pace literally changes the rules of the game and is predicated on the continued progress of several key areas. Though there are countless keys to the future, let's examine seven that are vital to sustain progress in delivering continued value through IT.

Storage

At the top of the list of the keys to the digital era is storage. Storage was once an unglamorous, but essential, component of every organization's IT infrastructure. Today, data storage has started to sizzle and is moving towards center stage with the spread of the Internet, multimedia, and new digital applications. SANs now give storage a universal appeal not previously realized by dedicated storage subsystems. Requirements for storage accompanied with effective storage resource management are soaring. Magnetic storage capacities are expected to continue to grow at 60% or more annually for the next 5+ years. Given this well-observed trend, the amount of available storage capacity should not pose limitations on any anticipated applications. The price of homes, cars, and clothing goes up each year. This is not the case with storage or other parts of the IT business. Storage prices are expected to fall at 30% or more annually during this timeframe making new, data intensive applications even more affordable. Experimental storage technologies such as Atomic Force Microscopy and combinations of optical and magnetic recording offer even more potential, though progress remains slow in some cases.

Semiconductors

Semiconductor advancement is vital to IT progression. Obviously, it won't serve any purpose to store unlimited amounts of data if we can't process it. The force driving the semiconductor business has now shifted from PCs to communications and networking. More people are connecting computing equipment to an ever expanding integrated and wired network of voice, data, and video links. By cramming more circuits on memory or microprocessor chips, computing is no longer the exclusive domain of large companies. We anxiously await the overdue arrival of nonvolatile RAM technology, possibly in the form of TMJ RAM (Tunneling Magnetic Junction Random Access Memory). Typical DRAM chips store 64256 million bits and microprocessors will soon be over 1,000 times faster than the largest mainframe computer was in 1986. Significantly lower microprocessor costs now enable computing power to reside in fabric switches and storage devices, key to the intelligent SAN. Microprocessors, the chips that process information, are headed for another boost in performance with the arrival of copper-based chips. Copper, as opposed to traditional aluminum based chips, is smaller and consumes less power. Semiconductor chips are expected to stay ahead of processor performance demands that the data storage explosion drives.

Software

Simply stated, software makes the computer work. Software development times must shrink. Key to achieving this is object-oriented programming that breaks computer programs into small reusable packages called objects. Reusing software cuts development time and creates a library of building blocks that other programs can readily use. The benefits of systematic software re-use have been appreciated for a long time but have encountered the human characteristic called "resistance to change." Average software development practice falls far short of what is theoretically possible and improvement has been minimal because of this resistance. The Information Age will be held back without fundamental changes occurring in software development. Specifically, the biggest challenge SANs face is the timely availability of SAN management software. Looking a few years ahead, we can imagine a more sophisticated developer's workbench taking advantage of the existence of plentiful libraries of re-usable components. The great majority of developers will assemble software components into finished systems, much as server manufacturers assemble finished systems from chips and motherboards. They will have much more time and energy to spend on analyzing requirements and making sure the end users are satisfied. The inherent flexibility of a component-based system makes it much easier and quicker to change the software as business needs evolve.

Parallelism 

Parallelism offers an architectural multiplier for performance, bandwidth, and capacity and is expected to become a more widespread critical future architecture as we deal with larger computational problems. Performance boosts come from connecting multiple processors together in parallel inside a single box. When a request is generated, software breaks it up so that each processor completes a different part of the request in parallel, making the response time many times faster. We will continue to see more parallelism in data storage systems as the array concept will transcend both disk and certain tape applications. As capacity growth dwarfs device performance increases, I/O performance improvements are gained by striping data across (N) disks or (N) tape drives resulting in (N) times the data rate of a single drive as data is transferred in parallel. Parallelism will become a primary factor in scaling processor and storage subsystems.

Optoelectronics

Optoelectronics is now the backbone of the information age and is clearly headed for widespread acceptance. Fiber optic technologies provide the bandwidth for large data elements to be transmitted from geographically distant sources has accelerated the SAN initiative. Today, large organizations are experiencing increases of 65 percent or more in Internet/intranet message usage while data traffic exceeds 40 percent annual growth. This growth rate may actually prove to be small compared to the rate driven by bandwidth-intensive applications that are just beginning to appear and further the need to establish the fiber infrastructure. Copper-based transmission lacks the long distance capabilities of fiber, though presently competes favorably in performance with fiber. Later, fiber will offer performance levels that exceed the limits of copper. New technologies such as dense wave division multiplexing will increase the number of communications channels within a fiber optic cable, thereby boosting bandwidth without requiring new cable. The costs associated with installing fiber, however, remain a factor, but are falling. Longer term progress will hinge on the amount of available and affordable bandwidth trying to keep pace with demand.

Speech Recognition

Speech recognition represents one of the major technology discontinuities in our future as it allows natural language to be used to interact with information technology. Learning a specific programming language or syntax limits the use of technology to a select portion of the world's population. Speech recognition has made progress in the last few years with commercial dictation systems now readily available. However, this is only the beginning of the long-awaited march to make speech work uniformly well for a wide range of users. The storage demand generated by speech is not as great as other applications, as one hour of telephone quality recording generates roughly 22MB of storage. With a large lexicon of words (tens of thousands of words), meaning can often be ascertained only by contextual reference, making universal speech recognition several years away. The technological advances in microprocessors and improvements in mathematical models are making the task of speech recognition easier, if only by throwing more computing power at it. Speech recognition will be fundamental to taking IT to the general population.

Wireless Transmission

Finally, we must include wireless transmission as a key to the future. For years, wireless transmission has enabled deployment of radio and TV broadcast. Our mobile society now looks toward wireless communication for many of the same uses. Increasing amounts of data are now being sent via wireless from a variety of applications such as e-mail and web browsing. Development of the wireless WAN is being fueled by a variety of mobile computing and palm-type appliances. Mobile users potentially become nodes on a SAN storing and retrieving data. Wireless permits nearly anyone to participate in the information age, regardless of the communications infrastructure that may exist at his or her specific location. Environmental transmission issues and security remain an impediment in delivering consistent wireless service levels, though they are improving.