ICMR Case Studies and Management Resources
 Asia's Largest Online Collection of Management Case Studies

Case Studies | Case Study in Business, Management

Quick Search


www ICMR


Search

  

AMD Opteron

Ravi Madapati 
Faculty Member
Icfai Knowledge Center

<<Previous

Opteron

Opteron was designed to run existing 32-bit applications and offer customers a smooth transition to 64-bit computing. The processor provided a dramatic leap forward in performance. It also reduced the Total Cost of Ownership (TCO)[1]. Opteron was offered in three versions: The 100 series (1-way), the 200 series (1 to 2-way), and the 800 series (up to 8-way). It provided a scalable architecture that AMD claimed was designed to meet current and future business needs. The Opteron was designed to scale from one to eight processors. This aided system designers by reducing the cost and complexity of building servers and workstations. It also reduced cost and increased server scalability.

 

Hyper transport technology

One of the most important features of Opteron was the HyperTransport Technology. The technology was designed to increase overall performance by removing I/O bottlenecks[2], increasing bandwidth/speed, and reducing latency[3]. For workstation users, this meant increased graphics throughput (up to 8x AGP), quicker loading of applications and large data sets, better multi-tasking, and smoother transition between applications. HyperTransport technology was primarily targeted at the IT and Telecom industries, but it was useful for any application where high speed, low latency and scalability were necessary. AMD believed this technology reduced the number of buses while providing a high-performance link for PCs, workstation and servers, as well as numerous embedded applications and highly scalable multiprocessing systems.

Architecture

AMD’s technology was designed to allow customers to migrate to 64-bit computing without any significant sacrifice of the existing code base. The technology was designed to provide full speed support for x86 code base, offering high performance levels for existing 32-bit apps. It provided a large memory which was useful for computationally intensive applications, such as databases, ERP, decision support, scientific and technical modeling, etc. It also helped lower TCO and network management complexity through a unified architecture for desktop, notebook, workstation and server, and platform flexibility.

The Opteron’s customer segments included worldwide enterprises, small-medium businesses, and government/education institutions, companies who required faster database transactions, or support for more users on e-commerce type applications, customers needing quick graphics response such as CAD and DCC industries with computational intensive tasks for financial modeling and scientific applications.

Though Opteron was designed for high-end servers, it could also run like 32-bit (Pentium and Athlon) processors in most PCs. Opteron could run 32-bit Windows and Linux software, as well as future 64-bit code. A PC version of Opteron was also expected to be available unlike Intel’s Itanium 2. Opterons were priced from $283 to $794, while prices of Itanium 2 processors ranged from $1,338 to $4,226.

Opteron’s design made it fully backward compatible with existing 32-bit applications. That
differentiated it from Itanium 2, which used a different architecture. AMD believed that this crucial difference gave Opteron an edge over Itanium 2.

More>>.Page2

[1] Total Cost of Ownership (TCO) is a model developed by Gartner Group to analyze the direct and indirect costs of owning and using hardware and software. Managers of enterprise systems use various versions of TCO to lower costs while increasing the benefits of information technology deployments.

[2] Any socket in the back of a computer that you use to connect to another piece of hardware is called an I/O (Input/Output) port. CPU speeds are improving at a dramatic rate, while disk speeds are not. This technology shift suggests that many engineering and office applications may become so I/O-limited that they cannot benefit from further CPU improvements. This is called an I/O bottleneck.

[3] In networking, latency is the amount of time, it takes a packet to travel from source to destination. Together, latency and bandwidth define the speed and capacity of a network.

© Icfai Press.  Global CEO • December 2003 ,All Rights Reserved.


Copyright © 2008 ICMR Case Studies in Business, Management.
All rights reserved.
Terms of Use | Privacy Policy | FAQ