HP has offered 64-bit computing since the late 90s. Over the past 15 years it has expanded 64-bit computing throughout its enterprise computer line. We connected with Chip E. Greenlee, III, HP’s director of marketing and solutions for the financial services industries, to get the company’s view on why 64-bit has become mainstream for the financial industry’s business-critical and data-intensive applications.
Describe HP’s commitment to 64-bit computing.
HP is deeply committed to 64-bit computing throughout the enterprise. Our Integrity and ProLiant servers both feature full 64-bit capabilities with complete 64-bit operating environments as well as support for 64-bit applications.
64-bit computing was first introduced into HP's HP9000 (PA-RISC based) systems and Compaq's Alpha-based systems in the later 1990's. Both companies invested with the leading enterprise software vendors like Microsoft, Oracle and SAP and with many financial services software vendors to bring the benefits of a 64-bit environment to our customers. The numbers of applications that benefit from 64-bit capacities continue to grow.
Nearly all of HP's current line of enterprise computers can support 64-bit computing and HP continues to invest with independent software vendors to take advantage of 64-bit computing.
What is driving the need for 64-bit computing?
Primarily it is capacity that drives 64-bit computing. This is nothing new. Disk storage was the first part of the data center that pushed past the 32-bit boundary. More than 4 gigabytes (4 billion) of capacity has long been needed in storage systems and now widely needed in computer systems to manage ever-growing data demands.
The ever-dropping price and increased capacity of main memory and disk storage is also a key driving factor. Installing more than 4 gigabytes of memory was once impossible but is now common in our larger systems. HP's Integrity Superdome servers have a capacity of 1 terabyte - 1000 gigabytes of memory! And this is doubling every 12-18 months. Customers do the ROI calculations of these larger capacity systems and quickly find business value in the investment.
This growth in data is also fueled by the ever-growing demands of the customer. Tracking financial transactions for a subset of the business might fit in a modest amount of storage; examining the entire enterprise on a moment-to-moment basis requires much greater capacity. Most current systems don't have the capacity for all the data that a financial enterprise might want to collect and analyze so ever-larger capability systems will be needed.
Finally, virtualization is a key technology to reduce the costs in today's data centers that utilizes the greater capacity in 64-bit systems. This capability to virtualize operating environments better utilizes memory and allows the consolidation of smaller systems into the more efficient, higher capacity 64-bit systems, saving not only in hardware and software costs but also in footprint, cooling and other ancillary costs.
What do you see as its biggest advantages?
Simply put, 64-bit computing enables the customer to solve problems that can't be done on 32-bit systems. 64-bit systems also improve efficiency by enabling applications to share the larger capacity among the many applications. These systems are perfect platforms to consolidate legacy systems to greatly reduce the number of systems in a data center. As an analogy, think how the airline industry made huge advances by having the capacity --- range and seats --- to cross oceans. Overseas travel before 747's was brutal.
HP invests significantly to improve reliability and availability capabilities in our 64-bit systems since they are often used in business- critical environments. Large-scale systems are not simply HW with extra capacity; instead they are systems built from the ground up to be more reliable than a simple 1 or 2 socket system. For example, these systems can have 128 DIMM memory slots, which are more reliable and resistant to failure than the 4 or so sockets in small scale servers. Customers can confidently deploy our systems and gain the capacity advantage of 64-bit computing and the reliability characteristics of mission-critical environments.
For customers who have extensive application suites designed for 32-bit platforms, will it be a difficult or costly transition?
No. Many older applications can remain 32-bit and never migrate. They often benefit from the higher capacity systems since they may link to databases and other applications that have been migrated. Some environments like Microsoft's .NET environment have been designed to have the nearly transparent ability to move forward without costly migrations.
For some applications, migration to a higher capacity environment will require rework but our experience with our ISV partners over the last several years has been that this rework is minimal in return for the performance and throughput advantages.
In terms of the strategic future of 64-bit computing, when do you see it moving into the mainstream?
It's already happened! Wherever higher capacity generates business value, migration will occur. In the financial services sector many of our customers have migrated to 64–bit platforms for business-critical and data-intensive applications including on-line banking and brokerage, customer analytics, portfolio management, mutual funds accounting and many compliance-related applications.
How does 64-bit computing stack up to the mainframe?
64-bit processors can provide the capacity of mainframe systems, but it is the system vendor that integrates those processors into systems that determine how well they compete with mainframes. Our 64-bit systems, for example, have all the critical capabilities of the mainframe --- reliability, availability, and scalability --- without the high cost. Building systems with industry standard architectures (x86 and Itanium) and adding value around the processor allow us to provide all the advantages of the mainframe but on an industry standard cost curve.
Are there any barriers that still exist in terms of competing with the mainframe?
Not really. Any migration must be evaluated from an ROI standpoint. The high cost of mainframe computing often enables one year or less paybacks in migration to 64-bit. With the high capacity and mainframe capabilities, 64-bit systems can take on any of the roles historically filled by mainframes.
In what ways will users be able to take advantage of the benefits of 64-bit computing from the desktop?
Many desktop applications are not capacity limited and don't need 64-bit computing. Desktops can often connect to higher capacity systems and indirectly benefit. As desktop machines locally operate on more data, the local operating system will be able to utilize larger 64-bit memory capacity. The financial industry often does have more capacity-limited desktop applications than general IT in areas like trading workstations, financial modeling, etc.
For more information, please visit
http://www.hp.com
