|
Benchmark
Ziff-Davis Winbench* 99 and Intel® Application Accelerator Boot Tester. Performance tests and ratings are measured using specific computer systems and/or components and reflect the approximate performance of Intel products as measured by those tests. Any difference in system hardware or software design or configuration may affect actual performance. Buyers should consult other sources of information to evaluate the performance of systems or components they are considering purchasing.
Boot / Operating System Up - Time Methodology
Overview - Any measurement of an operating systems boot can be measured by an exact process or methodology. However, variation is almost always present in the operation systems (OS) data structures which can affect outcomes of any boot processes or operating systems up-time calculations. Therefore, high population numbers (N>30) and statistical variation measurements are equally important to the results since methodologies should be made repeatable.
Also, it is important to understand that each computer system is unique and your results may differ than those published based on similar methods.
Accelerated Disk Input/Output - Intel Application Accelerator has the ability to substantially increase the storage subsystem speed by increasing the data rate, from the disk to the system, using special software technologies. Consequently, all boot processes, except for the BIOS, are inherently accelerated.
General Methodology - The following steps identify some basic ways to measure a computer systems boot model or an operating systems up-time.
Generally the total system boot time can be obtained from the RDTSC register on the possessor (time stamp counter). This register is set to 0 at complete power off (or S5 state). The value held is in clock-ticks which can be translated into seconds based on the processor frequency.
Additionally, the Microsoft* Windows* OS uptime can be obtained by calling the function GetTickCount as declared in WinBase.h. The DWORD returned is in milliseconds which can be translated into seconds to give a picture of the total operating system up-time.
- Identify a point on the desktop for reference: It is appropriate to identify a point at which the desktop is reached for the total process to end. Examples of different end points are:
- appearance of the desktop icons
- after all the taskbar applications have started
- after all the resident services have started
- the ability to start a program from the desktop
Initialize a timer: To measure a process, a timer must be present to interpret the event. Start the timer at the beginning of the boot process; either from complete power off (RDTSC) or at the beginning of OS load.
Stop the timer: When the boot process has ended and the identified point on the desktop reached, the timer is stopped.
Calculate the difference: This produces the result.
Methodology for measurements of Intel Application Accelerator on I/O intensive applications
Overview - Every software application has to, at some point, access the computers hard disk for data needed to launch, run, and exit. This sequence is known as a 'Usage Model.' All applications do disk intensive reads and writes (disk Input/Output {I/O}) at different points in the usage model scheme. For instance, graphics based apps tend to load most of the application data at launch, while some utilities access the disk regularly throughout the usage model.
Generally, the data is delivered to the system from the disk at a finite rate; hence representing a measurable process. While no single numerical measurement can completely describe the performance of a complex system like usage models, trends can be derived leading to important conclusions about system performance.
Accelerated Disk I/O - Intel Application Accelerator has the ability to substantially increase the storage subsystem speed by increasing the data rate, from the disk to the system, using special software technologies. Consequently, all disk I/O operations within software applications are inherently accelerated, resulting in quicker application usage models.
General Methodology - The following steps outline one way to measure an applications usage model. An example software application could be game such as Quake* III Team Arena. To measure disk I/O you could measure the process of starting the application, and loading the texture and level. It then breaks out of the game before being passed to the video card. This represents game disk I/O such that the level, game data, and textures are all disk based and are loaded into memory at the start of a match; a usage model process which can be measured.
- Identify what is to be measured: Pick the application operation that represents disk I/O. In this case, "loading a level" is the disk I/O component of the application.
- Initialize a timer: To measure a process, a timer must be present to interpret the event. Start the timer.
- Activate the process: Once the timer has started, activate and run the application operation.
- Stop the timer: When the application operation process ends, the timer is stopped.
- Calculate the difference: This produces the result.
* Other names and brands are the property of their respective owners.
For more information on performance tests and on the performance of Intel products, reference http://www.intel.com/performance/index.htm or call (U.S.) 1-800-628-8686 or 1-916-356-3104.
This applies to:
|
