The $US5.8 billion server market in Asia/Pacific (excluding Japan) in 2002 is expected to be the fastest growing region in the world through 2007, growing at a CAGR of 8 per cent by revenue and 19 per cent by unit shipments. The Peoples Republic of China, Korea, Australia, Taiwan and India will be the leading markets underpinning this growth. The actual drivers will vary by country, for instance the server market in China and India will be driven primarily by basic IT infrastructure and e-enablement spending, while Australia, Korea and Taiwan will have capacity expansion, new applications, and consolidation trends as the primary drivers. However, there are certain pivotal trends that will cut across all markets and touch all users. Following are certain key trends that IDC believes will impact the server market around the region.
x86-32 servers come of age: The increasing performance of x86-32 bit servers along with the accompanied robustness of the operating environments will drive their adoption in the enterprise space, a segment that has eluded them in the past. The reliability, availability and scalability (RAS) capabilities, so important for these applications, have been improving on x86 servers and this has helped in increasing their penetration in the business processing, business intelligence and lately high-performance computing markets. The increasing capabilities have led some users to consider them as a platform for consolidation as well.
RISC-Unix servers increase penetration in the mission-critical space: RISC Unix servers are expected to continue dominating the mission-critical space. Constant RAS improvements and availability of better virtualisation and integration tools are helping users do a lot more with these systems. Over the last many years, vendors have been able to offer many mainframe-like capabilities on these systems, which have encouraged users to consider them for workloads that were largely the domain of mainframes in the past. However, due to increasing competition in this space one can expect to see further price-erosion for vendors in this space over the next few years.
EPIC architecture will drive Windows and Linux into enterprise space: EPIC architecture should begin to gain momentum in 2004, driven by HP Unix, Windows 2003 and 64-bit Linux versions. The architecture will make a special difference to Windows and Linux-based server platforms by increasing their penetration in the enterprise application space. The hardware and software limitations, that have impeded the growth of these operating environments into areas that demand 64-bit readiness and higher RAS, will be resolved by EPIC to a certain extent.
Consolidation will gain momentum driving integration and utility computing over time: Customers in the Asia/Pacific region are opening up to consolidation at all levels of their infrastructure. While there is still more of physical consolidation, we should soon see application-level consolidation as well. The finance industry is by far the leading segment here but telecomms, manufacturing and distribution companies are embarking on this as well. We believe this will in time lead to greater application integration within organisations using Web services, and will facilitate models such as utility computing.
Utility computing adoption in the region will lag well behind US: Users in the Asia/Pacific region will need far greater convincing and will probably expect to see some successful models in their markets before adopting utility computing. Much depends on how successfully these models are implemented by vendors providing such services and how seamlessly users are able to continue their business with the same or greater effectiveness than before. There will be apprehension in letting go of core applications, and hence we will see the adoption in phases, as and when it occurs. Consolidation, mentioned above, can be a key driver of this model as users who consolidate any given application or service may see sound reason in adopting utility-based computing as the next stage. This could also work the other way as users could look at utility computing as an easy way to consolidate certain types of applications, such as those that demand fluctuating computing-resources over a period of time.
There are bound to be changes moving forward in the way users are adopting servers today, since there is an increasing consciousness to optimise cost, apart from increasing organisational efficiency. We will see over time that users will become more open to adopting new technology, provided vendors are able to deliver to commitments.
Avneesh Saxena is VP, Asia/Pacific computing systems research, IDC