Q1) In summer 1998, what is the position of the Enterprise Server Group (ESG) in its industry? How has it evolved? Why? Ans. Davis was the general manager for Enterprise Server Group (ESG) (Exhibit 1 & 2). In summer 1998, HP had to decide whether it should recommend a large share of its outlay to its Performance Architecture (PA) that is UNIX based or should shift its resources to the development of a new chip that would hold the future of enterprise computing. ESG was formed in 1997. The creation of ESG was welcomed by companies like IBM, Hewlett Packard, Compaq and SUN.
It was responsible for producing scalable, high performance computing systems and it ran complex applications helping customers run their businesses. ESG’s products were built around RISC microprocessors and UNIX operating systems which led to a tremendous increase in the performance of the computing systems. By 1997, ESG had an installed base of more than 1. 3 million and had sales of almost 10B. HP’s net revenue and net sales kept increasing from 1992 to 1997 (Exhibit 3 & 4). HP had maintained number one position in UNIX server vendor, by sales, since 1993, through its HP-UX products.
In 1994, as HP joined hands with Intel to develop the IA-64 architecture, its competitor, SUN started a campaign that HP was not committed to RISC-UNIX systems. This had eroded HP’s customer base. The current situation was critical because any reduction in the resources allocated to the Performance Architecture (PA) would lead to erosion of market share and revenue. Q2) Why did HP get involved in developing the IA-64 architecture? Ans. ESG’s products were built around RISC microprocessors and UNIX operating systems which were increasingly powerful and reliable.
SUN, IBM and HP were some of the known makers of RISC- based computer systems. On the other hand, Intel Pentium Pro Microprocessors and Windows NT had combined into a powerful WINTEL system and were capturing the market share rapidly because of their low cost. Technologies often improve faster than customer requirements demand. This enables low-end technologies to eventually meet the needs of the mass market. Thus, if the low-end market is neglected, it can become a breeding ground for powerful competitors. HP, SUN, INTEL and ther powerful companies faced the same consequences. In 1997, UNIX based systems continued to dominate the market with 37. 3% share (Exhibit 6) and accounted for revenues of nearly $25B (Exhibit 6). WINTEL accounted for only 9. 3% market share (Exhibit 6), but sales of products based on WINTEL grew 66% from 1996 to 1997 (Exhibit 6) and this trend was intended to continue.
Although UNIX was expected to be the prime operating system in the market, WINTEL was rapidly progressing. The growth rate of UNIX-based systems from 1996 to 1997 was 9. % and was expected to slow down from 1997 to 2001 because of WINTEL-based systems. Technology S-curves indicate that technology improves slowly at first because it is poorly understood, then accelerates as understanding increases and then tapers off as approaches limits. Many companies believed that RISC architecture was approaching the limits of its performance. While WINTEL had captured the low end market, the cost of designing RISC chips were increasing and software application development had become more costly.
Due to these reasons Intel and HP collaborated on the IA-64 architecture, which was the basis for Intel’s Merced processor and other future 64-bit chips for high-end servers and workstations. The goal of the IA-64 architecture was to dominate the high-end market (Exhibit 5) and improve on today’s RISC and 32-bit architectures. This suggests that the development of IA-64 was an incremental change for Intel and an architectural change for HP and Intel. Merced was the first of a family of processors that would enhance the Intel architecture with 64-bit capabilities.
IA-64 offered full binary compatibility with Intel’s IA-32 as well as HP’s PA-RISC processors. According to Intel and HP, in the long run Merced and follow-on IA-64-based processors would have their largest impact on Information Systems, due to advances in chip technology, known as Explicitly Parallel Instruction Computing (EPIC). These advances would give IA-64 clear differentiation from RISC-based systems by more effectively using parallelism in processing – allowing users to apply IA-64-based systems more effectively on specific high-end applications.