|  AMD Business Home  |  AMD.uk  |  Support and Drivers Go Search
Looking for AMD Server Benchmarks?
We have our AMD Opteron™ Processor-based Server Benchmarks in our products section.

Real Benchmark 

Latest News 

Many organizations rely on benchmark scores when making computer buying decisions. Benchmark scores might measure the raw speed at which computers perform certain tasks, but they don’t necessarily demonstrate how long it takes real users to complete real-world workloads. As a result, organizations may end up significantly overpaying for a system for no significant gain in productivity. The video to the right illustrates this point by comparing productivity of two popular desktop processors – Intel Core i5 650 and AMD Phenom II B55.

AMD has recently resigned from the BAPCo organization and withdrawn the support for SYSmark 2012. The latest developments on this topic can be found here:

Coverage and Commentary

Important Links:

Resource Center 

Did you know that the foremost thing that modern organizations need from their computer systems is Productivity?

Did you know that organizations can overpay by as much as $196 more per system1 for a non-AMD system with a comparable level of productivity?

This means there is a significant amount of budget waste for medium to large enterprises and public sector agencies that buy hundreds or thousands of systems at a time. That money could be invested in other areas or even saved.

When it comes to purchasing a PC, buyers must take into account a variety of factors including performance, investment protection, security, management, reliability, ease of deployment and price. While performance is a key factor it can be the most elusive to measure.

AMD recommends that customers buy the right system based on their real-world needs. A typical business user can be classified as Basic/Task user, Mainstream/Knowledge user and Power user based on their usage requirements.

You can find more details on the Price of Productivity and the recommended AMD systems for the 3 tiers of business users are in the Benchmark Brochure.

To help customers in their purchasing decisions, AMD has created a multi-tasking script to simulate real-world usage of common productivity applications. You can download the details of the mutli-tasking script and the testing instructions from the Test Yourself tab.

You can learn more about today’s benchmarks for the commercial environment in the AMD Benchmark Whitepaper.

Test Yourself 

AMD has created a multi-tasking script to simulate real-world usage of common productivity applications. You can test the script yourself to measure productivity on your system by simulating multi-tasking activities that are typically performed in a business environment. Here are the various components of the test package:

  • 1) How to run the multi-task script execution guide: This guide helps you to run the multi-tasking script to simulate real-world usage of common office productivity applications.

  • 2) Multi-tasking demo files: This zip package contains all the demo files that are part of the multi-tasking package. Follow the instructions in the guide on how to use these files for your test.

  • 3) Multi-tasking script code: This is an automation script for Windows GUI used by the execution module (Item 4). Place this script in the same folder as the execution module. To know more about what is an automation script, click here. If you would like add or remove the commercial applications or change the order of execution, you can modify the script yourself and re-compile into an executable file.

  • 4) Multi-tasking script execution module: This zip file contains the executable that runs the multi-tasking script. Unzip this file in the same folder as the other files and follow the instructions in the guide on how to run this application.

    Note: Please read the instructions in the execution guide carefully. Although the script is simple, attention to detail is necessary for accurate results.

To change the source code to create your own multi-tasking script with different set of applications, download and install the AutoIt compiler.

Following installation, simply open the source code and build your custom script. If you have the compiler and source code loaded on the actual test system, you could select, “Run” and the code will compile and execute.

Note: The source code is simply providing a series of keystrokes and mouse clicks. If you do not have the correct version of all of the software installed or the applications pinned to the taskbar in the right locations, the script will not run as designed.

If you have any questions on the execution of the script, please email the AMD Real Benchmark Team at real.benchmark@amd.com.

 

 


Misleading Benchmarks

SYSmark 2007, as well as SYSmark 2012, mainly focus on CPU intensive applications and virtually ignore the common GPU accelerated applications used in the business environment today.
 

IT Customers May Be Overspending

Learn more about making a PC purchase decision based on the time it takes to execute real-world business workloads.
 

AMD Technical Tools Available To You

See the difference in productivity yourself with tools that simulate real-world usage of common productivity applications.
  1. See AMD Performance Labs’ demonstration of AMD and Intel-based systems running a script of common productivity applications: http://sites.amd.com/uk/business/it-solutions/benchmark/Pages/real-benchmark.aspx#/45478 . In the demonstration, the AMD Phenom™ II X2-based system completed the script in 134.50 seconds while the Intel Core i5-based system took 133.98 seconds, a difference of only 0.52 seconds. According to pricing as of 12/1/2010 the AMD-based system depicted in the demonstration costs $600 (HP.com), while the Intel-based system costs $800 (Dell.com).
Value, Industry Standard Manageability and Security
AMD VISION Pro based enterprise-class desktops designed to deliver results:
Commentary and Coverage
“…BAPCo and SYSmark became obsolete as soon as AMD decided to take a different path.”
“…for something calling itself 2012 to include mostly 2010 applications feels out of place...”