Professional Soft Collection 1.02

home *** CD-ROM | disk | FTP | other *** search

/ Professional Soft Collection 1.02 / Professional Soft Collection 1.02.iso / test.all / speed.doc < prev next >

Wrap

Text File | 1993-03-05 | 18KB | 329 lines

▄ ▄█ ▀▄ ▄███ ▀▄ ▄█████ ▀▄ ╖ ╓─┐ ╥─┐ ╥─┐ ╓─┬─┐ ╓─┐ ╥─┐ ╥┌─ ▄▀██████ ▄▀▄ ║ ╟─┤ ║ │ ║ │ ║ │ │ ╟─┤ ╟┬┘ ╟┴┐ ▄██ ▀████ ▄██ ▀▄ ╙─┘ ╜ └ ╨ ┴ ╨─┘ ╨ ┴ ╜ └ ╨└ ╨ └ ▄████ ▀██ ▄████ ▀▄ ▄██████ ▀▄██████ ▀▄ S y s t e m S p e e d T e s t ▀█████ ▄▀ ▀█████ ▄▀ ▀███ ▄▀ ▀███ ▄▀ V e r s i o n 6.00 ▀█ ▄▀ ▀█ ▄▀ ▀ ▀ Copyright (C) 1993 Landmark Research International Corporation All Rights Reserved Valid command-line switches for this program are as follows: /b - bypass this initial screen /nv - don't perform initial video test /q - run in quiet mode (no sound) /xx - run for 'xx' seconds, then exit USER INTERFACE You may control this program with a mouse and/or keyboard. To scroll the text in this box, click on the up and down arrows in the text box border, or press the up/down arrows, or page up/down keys. To change the scale of the speed bars (normally yellow), click the left or right mouse button while the mouse cursor is on top of the bar. Left button increases the scale to make the bar shorter (violet), and right button decreases the scale to make the bar longer (cyan, or light blue). Press F3 and F4 to increase/decrease the CPU and FPU scales, or F5 and F6 to change the video scale. If you are running on a monochrome system, you will not be able to discern what CPU type or speed scales are active from a distance because color is used for those purposes. Press F1 for help, F9 to recalculate video speed, F10 to toggle the beeper on and off, and Esc to exit the program. You may run the program with one or more command line parameters separated by spaces to affect operation: SPEED /B - bypass initial help screen SPEED /NV - no video (disable video test) SPEED /Q - quiet (disable beeper) SPEED /40 - run for 40 seconds and exit SPEED /B /NV /Q /40 - all of the above THE TESTS The system speed tests tell you the most important information about the computing performance of your system: the effective speeds of the central processor unit (CPU), floating point unit (FPU, math coprocessor), and video display circuitry. Although hard disk access and data transfer rate affect system performance, we excluded those tests because they do not affect actual computation speed. CPU/FPU Tests The CPU and FPU speeds show performance relative to an IBM AT with a 6 MHz 80287 math coprocessor chip installed. For example, if the display shows CPU speed to be 22.53 MHz, it means an IBM AT would have to run at 22.53 MHz to have the same computing performance as this computer. If the display shows FPU speed to be 40.25 MHz, it means an IBM AT would have to run its 80287 math chip at 40.25 MHz to have the same floating point math calculation performance. The speeds are NOT the actual CPU/FPU clock speeds. CPU Test The CPU test forces the processor to show the effect of instruction cache. It is thus more accurate than previous versions. On color systems, the background display is gold for 8088 and 8086 CPUs, green for 80286, blue for 80386, and red for 80486. This allows you to see from a distance what kind of CPU is running the system. The display also shows in text the type of CPU installed. FPU Test The FPU test shows a value only if a math chip is installed or built into the CPU. Where possible, it shows the manufacturer of the chip (e.g., Intel, Cyrix, IIT). Note that the 80486 has an internal FPU, but other CPUs do not. The test accurately reflects the higher performance of a closely coupled CPU-to-FPU architecture. Screen Updates The program runs both CPU and FPU tests continually, and updates the display about 4 times per second. The speed will drop if you run another application concurrently or change CPU clock speed. You will be able to see the system speed degradation caused by multitasking or background network accessing, or by taking the sys tem in and out of turbo mode. Clock Speed / Multitasking Multitasking does not affect actual CPU clock speed, but allows less time for the speed test to run; turbo mode actually changes the CPU clock speed, and thus affects the performance of both multitasking and the speed test. Some systems do not have a turbo mode, but those that do usually allow you to invoke it with a front panel switch or the keyboard (normally Ctrl-Alt-plus or Ctrl-Alt-minus). Variable Readings As the test runs, you may notice the CPU and FPU MHz values changing slightly. This does not mean your system clock is running erratically, but rather that test results are being rounded off, and occasional interrupts in the background are using CPU time. If you want the fastest possible readings, you should boot DOS on the computer with no drivers or resident programs installed. Such software can cause the CPU to work on special tasks in the background while the speed test runs in the foreground. The special tasks will take CPU time and cause the speed test to give lower readings than it should. Large temperature changes can cause PCs to vary in performance characteristics. As a result your computer may run faster when it is cold than when it is warmed up. Video Test The video test checks pure data transfer rate from the CPU into video memory. It thus shows the maximum speed at which data can be read and displayed. The video test detects most popular video adapter chipset types, and displays the type and manufacturer on the screen. If it cannot determine the manufacturer, it shows just the generic type, such as monochrome, CGA, EGA, or VGA. Video / System Design The video test results are affected by video adapter design, CPU speed, and system board design. For example, some systems have fast I/O bus speeds and allow a fast data transfers to video adapters. Others, even some with very fast CPU speeds, have slow I/O speeds and thus slower display rates. Much of the power of a fast video adapter can be wasted if it is used in a slow system. On the other hand, a slow video adapter can impede the faster systems, giving diminished performance in display-intensive applications. Video Runs Once The program runs the video speed test only once, at the beginning. This is partly for aesthetic reasons, since it makes the display look unattractive. You may re-run the video test by pressing F9 (recalculate). You will need to do this if you take the system in or out of turbo mode while the normal speed test is running, because video speed can be affected by CPU speed. Video BIOS The video test does not show the effect of BIOS on display speed because most popular high-performance applications write directly to video memory, and do not use BIOS. Furthermore, most super VGA adapters come with a program that loads display BIOS into system RAM for maximum execution speed. SCALES All of the test results are shown on variable scales. The default scale values are different for each CPU type, but the default speed bar is always yellow. This allows the program to show how the performance of your system compares to the scale limits for your processor type. Thus, if you view the screen from a distance, the program accurately renders performance information without the exaggerated effect of display wrap-around or logarithmic scaling used in earlier versions of the program. If you see a yellow speed bar, you are seeing the default scaling for your CPU type. If you are running on a computer that is slow for the class of CPU it uses, the speed bars will be short. If the computer is fast, the speed bars will be longer. If the bar is too short or too long for your liking, you may change its length by adjusting the scale. However, when you change the scale, you will also change the color of the speed bars to either cyan or violet. This will allow distant viewers to recognize that the system is not using the default scale and that the length of the bars have a different meaning. If the bar is too long on a scale, you have a very fast computer. To shorten it and turn the bar violet, press F3 or click the left mouse button while the cursor is on top of the bar. F3 expands CPU and FPU scales, while F5 expands the video scale. If the bar is too short on a scale, you are running on a relatively slow computer. To lengthen the bar and turn it cyan, press F4 or click the right mouse button while the cursor is on top of the bar. F4 shrinks CPU and FPU scales, while F6 shrinks the video scale. HISTORY AND STANDARDS From 1984 through early 1993 the original Landmark CPU Speed Test (tm) gradually gained prominence and became the de facto world standard for measuring PC computing speed. There are several reasons for its rise to popularity: 1. Standard - it compared the speed of the computer it ran on to the speed of the first and only de facto standards that ever existed for personal computers: the IBM PC and the IBM AT. IBM never changed the architecture of either computer, but replaced them with the PS/2 line in 1987 and took the PC and AT out of production. Since IBM will never change the PC or AT, those will always be stable standards of comparison for PC performance, and Landmark will be the performance measurement standard. The original PC used an Intel 8088 CPU and optional 8087 FPU running at 4.77 MHz. The original AT used an Intel 80286 CPU running at 6 MHz, and a later version ran at 8 MHz; in both versions the optional 80287 math chip ran at 6 MHz. 2. Visibility - the Landmark speed bar that reached horizontally from left to right across the screen was visible and relatively readable from a distance of 15 to 30 meters (50 to 100 feet). This made it ideal for dealer showrooms and trade shows, so manufacturers and resellers enjoyed using it to show off their computer's speed. 3. Credibility - it was the most accurate benchmark test on the market because it considered a variety of factors that other programs did not, and it showed performance relative to the PC and AT. Other benchmark programs gave unreliable or misleading information: Norton System Index (SI) was too simple and overrated AT performance relative to the PC. MIPS (millions of instructions per second) used an instruction mix that was somewhat arbitrary. Whetstone, a test of general computing combined with floating point or scientific math calculations, was originally written in FORTRAN, so the results were badly skewed by both compiler efficiency and the ratio of floating point math instructions to arithmetic and loop instructions. Dhrystone, a test of general computing performance, was originally developed to test ADA compiler efficiency, but was converted to the C language for execution on the PC. Because of the plethora of C compilers in use, it suffered from the same kind of skewing as Whetstone. 4. Results - The speed test displayed its results graphically with fixed scaling, and it beeped while it gave continuous screen updates. All the other programs, such as SI, MIPS, Power Meter, CheckIt, and QA Plus, ran through once and stopped. Thus, in real time it showed how much system performance degraded during network accesses or other multitasking program activities. Furthermore, the instant update gave immediate feedback to users, and gratified them that the system was actively running. Finally, benchmark graphic displays in QA Plus and CheckIt automatically scaled test results so that their speed bars had no meaning when viewed from a distance. ABOUT LANDMARK Landmark Research International Corporation is a high-technology research, development, manufacturing, marketing, and sales organization with headquarters in Clearwater, Florida. The original company, Landmark Software, was founded in 1984 in Sunnyvale, California, to produce the Landmark CPU Speed Test, Landmark Setup program, and assorted system utility programs. In early 1990, the company was restructured with expanded business goals under new leadership and given a new base of international operations on Florida's west coast. Landmark develops and supplies computer-related products to many types of customers: dealers, distributors, computer manufacturers, 3rd and 4th party service companies, corporate micro management and repair centers, and software/hardware developers. LANDMARK PRODUCTS Existing products include the following: 1. The PC Probe (tm) advanced testing program for PCs - it features replete system information, complete diagnostics, a full suite of benchmark tests based on the Landmark System Speed Test, CMOS RAM setup and low-level hard disk format utilities, remote operation, and the ability to display actual photographs of the circuit or component being tested. The complete and perfect program for end users, technicians, dealers, and OEMs to find out what's in the computer and how well it works. 2. The AlignIt (tm) maintenance kit for 3.5 and 5.25-inch floppy drives of all popular densities - it includes a pre-lubricated cleaning diskette guaranteed for 180 10-second uses, and a patented diagnostic and alignment spiral diskette that tests and allows drive alignment to within .3 mils. Perfect for technicians and end users alike to avoid costly drive repairs. 3. The KickStart (tm) family of diagnostic cards for PCs - the cards assist technicians in identifying failures in systems that won't boot, and help Value-Added Resellers to run remote diagnostics on client's computers. LEDs or digital displays show failing power-on self-test numbers during boot. Built-in circuitry on some models provides serial, parallel, floppy, and hard-disk I/O, as well as special test hardware for motherboards. The design engineer's and test/repair technician's constant companion. 4. The Burn (tm) advanced manufacturer's burn-in test for PCs - it features menu or command line operation via local console or remote terminal, nested count-based or time-based testing loops, pre-configured user-modifiable test sequences, and error logging to printer, file, or terminal. Site licenses available to PC system and motherboard manufacturers. 5. The Memory Boss (tm) family of Expanded Memory Managers - fully compatible with Lotus-Intel-Microsoft Expanded Memory Specification 4.0, Memory Boss operates on XTs (disk-based emulator), ATs (extended memory based emulator), 386 ATs (virtual 86 mode based emulator), and Chips and Technologies NEAT chipset EMS circuitry. Gives Lotus, Windows, and other applications huge memory space to operate in. Perfect for all levels of users; OEM licenses available. 6. The JumpStart (tm) family of self-configuring motherboard BIOS for 80286, 386, and 486 ISA and EISA computers as well as XTs. Fully IBM-compatible and non-infringing. Available in ROM for system developers and troubleshooters. Source and object code (label) licenses available for manufacturers. Other products that have been in development and may be available include graphic image processing and file conversion; FAX/Data modems with voice mail software; automated test equipment cards for ATs; and software developer libraries for XT/AT/386 AT virtual memory management, graphic image data base management, and menu systems. Contact Landmark for more information. When you think of standards, think LANDMARK. WARRANTY DISCLAIMER The Author cannot and does not warrant that any functions contained in the Software will meet your requirements, or that its operations will be error free. The entire risk as to the Software performance or quality, or both, is solely with the user and not the Author. You assume responsibility for the selection of the program to achieve your intended results, and for the installation, use, and results obtained from the Software. The Author makes no warranty, either implied or expressed, including with- out limitation any warranty with respect to this Software documented here, its quality, performance, or fitness for a particular purpose. In no event shall the Author be liable to you for damages, whether direct or indirect, incidental, special, or consequential arising out the use of or any defect in the Software, even if the Author has been advised of the possibility of such damages, or for any claim by any other party. But if you print it out, light it on fire and your house burns down, don't blame me. All other warranties of any kind, either express or implied, including but not limited to the implied warranties of merchantability and fitness for a particular purpose, are expressly excluded.