Software guide: Useful Utilities for overclocking and not only. WORK WITH PHORONIX TEST SUITE, or how to test the performance of virtual machines in Linux? Linux Processor Testing Inscription

To properly test the processor cooling system and the stability of work in acceleration, when there is also a voltage to raise the voltage to achieve high frequencies, you need to load it under the string for quite a long time. This will help only special utilities that know how to load the processor on the rigging, complex, but meaningless calculations. Here the main thing is to take the "right" program to work effectively using complex floating point formulas to handle a large amount of data. Perhaps one of the best utilities created Intel, which is called LINPACK. True, it was done by programmers for advanced users, without bothering over windows with ruffles, limiting the command line. A black screen with incomprehensible running strings robs ordinary users of the computer into a stupor that do not understand what they do with these and where to click the mouse. Glory to God found good people who screwed a graphic shell to Linpack, calling their creation Linx.

LINX does not have to install, and you will need to download the archive file, unpack it, go to the created folder and start the Linx.exe executable file. To delete the program, it will be enough to delete the folder where the LINX files threw off, the main thing is not forget, as it is called and where to be.

The fact that intel is not worth worrying about the creation of tests are not worried about the efficiency of working with AMD processors, LINX perfectly loads processors as Intel and AMD.

After launching LINX, the program is actually ready to work for this enough to press the "Start" button at the top of the right side. Further, gain patience and wait until the specified number of tests in a circle takes place. Upon completion of each test cycle in the table, its results will be added, namely the time, the data processing speed, the number of streams used and the size of the matrix of the linear equation system, which was used for calculations. At a time, loading the processor to the maximum, allowing you to check it on the stability of the work and the efficiency of heat removal of the cooling system.

To stop the test run, just press "TOP".

In the main window in top string Under the main menu there are sections where users can change:

"Task volume:", indicate the size of the matrix of the system of linear equations, which will be used for calculations. The more, the longer the calculations go. Note, this parameter is associated with "Memory", that is, when one of them changes, another will automatically change.
"Memory (MIB)", volume random access memorywhich the program can use the program. As mentioned above, with an increase in volume, it will automatically change "the volume of the task:", the same way.
"Perform" how many test runs will be done in a circle.

Of the interesting features of the program, it costs to note the support of obtaining data on the temperature of the processor, the frequencies of the core and supply voltage from the EVEREST / SPEEDFAN applications and display them in the form of graphs during tests. Plus Allow stop tests when errors occur or if the processor is heated above the specified value (only when working in the Evered / Speedfan pair). Also in the settings, you can specify the number of threads, which will be performed at the same time, select the title of tests 32 or 64-bit, the priority of testing processes (it is better to leave the default, so that the computer does not slow down), and limit the scope of the program memory used, so that the operating system is not Brake.

To date, LINX, and if you are completely accurate, then linpack if you lean the graphics shell, the only test that allows you to adequately check the stability of the processor. Consider, this program creates very high loads on the processor, under which it will never work at home, and it is not desirable to constantly launch it on your computer for several hours. Especially in the light of the fact that everyone intel processors The last generations under the lid of the terminal, which, with long-term exposure, high temperatures, can over time to lose its properties.

The program works both in 32 and 64-bit operating systems. The program interface is translated into English and Russian, users must download the desired version Program with appropriate localization.

Question computer diagnostics PC or laptop owners are usually puzzled with regular errors in their work. But a considerable share of enthusiasts is enjoyed by overclocking processors, video cards, overestimation of working frequency frequencies, etc. Naturally, uninterrupted work in such emergency situations is not guaranteed, and overclockers in the process of overclocking have to check the stability of the operation of an accelerated device. Check processor - Mandatory stage when overclocking the CPU. The proposed article is just devoted to the issue of verification of the processor under the maximum load after overclocking. Utility LINX 0.6. With a very simple interface will help us in this.

Recently, on the site there was an article about which only mentioned the need to diagnose the CPU, but there was no explanation. Today I decided to fill this gap and describe the processor check procedure. For computer diagnostics It will take only one utility designed to maximum CPU load computing tasks. The name of this compact program is LINX. On the this moment relevant is " LINX 0.6.5." In addition to testing stability, the utility can be applied as a good benchmark.

From the user's point of view LINX 0.6. Very simple and convenient thing with a simple and intuitive interface (Russian and English options). It is supported by both Intel processors and AMD CPUs. In the settings, you can select 32/64-bit mode, the number of threads, use or turn off the Hyperthreading technology from Intel, set the number of runs, monitor the testing of the test "in live air»... automatic saving the specified processor check parameters It will allow testing with the same settings upon subsequent launches, for example, with gradual acceleration of the CPU.

The main program window interface contains only a few controls. First of all, we are interested in the "memory volume" / "Memory" parameters and the number of targets "the number of times" / "Run". For computer diagnostics It is recommended to select all available memory in the first parameter (ALL button near "Memory"), and "Number time" / "Run" to install an experienced way: run one run, and calculate the required number of startups to work for at least an hour. In the "Settings" / "Settings" menu, select 32/64 bits mode depending on the installed processor and version operating system. To start / stop the test, use the "Start" / "Stop" button ("Start" / "Stop"). During the test, the test run indicator will display the time that has passed since the start time, and when the last pass is completed, the inscription "completed without errors for MM MS C" (or "FINISHED WITHOT ERRORS IN MM M SS S" in the English interface version).

With long-term testing of the unstable system, errors may occur. Item "STOP ON ERROR" in the Settings LINX 0.6 will stop further verification of the processor. This may be quite useful when errors occur due to overheating processor. On time, the stopped test may in some cases save life Checked CPU. If after starting the utility you fixed the temperature close to critical or exceeding such, then this is an explicit signal to what is necessary replace cooler Or reduce overclocking.

Using LINX 0.6.5. as benchmarka The performance of the processor will be displayed in the graph "GFFS" / "GFLOPS" of the final table of the main program window. So, for Intel Core I7-3770K.which was overclocked up to 4.5 GHz, the result will be the figure in the area 110 ... 115 Gflplops Depending on the settings. After a few runs of the specified CPU, the speed is specified in the table for each passage and in the status bar there is information about peak (maximum) performance in the Gflplops.

By the way, if you wondered processor acceleration, then puzzled and the expediency of inclusion of technology Hyperthreading. For Intel CPU. Not in all tasks from HT there is an increase in speed, but it unders acceleration opportunities "Stone" included HT very often. Judging by numerous reviews, from Hyperthreading there is not the slightest increase in FPS in most games - here already "playability" more depends on the video card.

Benchmark is a measurement maximum performance Computer, which is expressed in conditional glasses. Thanks to this, you can compare productivity different computersor the same computer after, for example, overclocking or andndertalling.

Benchmark and stress test is not the same thing. And with a benchmark and at a stress test, the system gets a full load. But the main goal of benchmarking is an assessment of performance, and the main goal of a stress test is a check if the system can function on the limit of its workload, or determine this limit. Although, in fact, some programs combine both functions.

Benchmark can be performed the length of the system as a whole or for the individual components: for central processor, video cards, I / O systems.

Linux has several programs to estimate the performance of the central processor, for example: sysbench, stress-NG. and pHORONIX-TEST-SUITE. Of them stress-NG. First of all, performs the functions of stress test, but it displays the received metrics, so it is quite suitable for evaluating and comparing the performance of the system.

Benchmark in Sysbench

sysbench is a utility command line. It is designed to evaluate the performance of servers with highly loaded DBMS, but is also suitable for conducting the benchmarks of conventional systems.

Installation in Ubuntu, Linux Mint., Debian, Kali Linux:

Sudo Apt Install Sysbench

Built-in tests:

fileio. - File I / O Testing
cPU. - CPU performance testing
memory. - Testing the speed of memory functions
threads. - Testing the performance of the flow subsystem
mutex. - Mutex performance test

To start the performance test of the central processor:

Sysbench CPU Run.

Note How the program starts: at the beginning there is a test name, then the option (in the first example there is no), and then the command.

For the program installed two completions:

10000 operations with numbers
10 seconds

Depending on what the first comes, the program will complete its work or after 10,000 events or after 10 seconds.

Modern processors are very productive and if the program ended very quickly, then the data can be distorted. For example, when evaluating the performance of the processor plays a role, for example, trottling (frequency reset). Trottling begins due to overheating or TDP exceeding. These effects are observed only at long distances of the processor. If, for example, the test ended in a second and you got n.processed operations, this does not mean that the processor for 60 seconds will perform 60 * N. Operations, since it will drop frequencies due to overheating and outputting the heat being installed in TDP.

For longer test execution, options are used. -CPU-Max-Prime and -Time. The first installs maximum amount Operations performed, and the second is the maximum time of the benchmark. With the simultaneous use of options, priority has -Time.

Modern central processors are multi-core and multi-thread:

Default sysbench Runs into one thread. Therefore, if you want to use all the kernels of your processor, use the option. -Threads.. I have 6 physical and 12 logical cores of the central processor, so I will use the value 12 to work all processors.

When using option -CPU-Max-PrimeThe less time the program is completed, the productive central processor:

Sysbench CPU --Threads \u003d 12 --cpu-max-Prime \u003d 100000 Run

The program completed the work too quickly - in 10 seconds it is unlikely that the processor managed to expose serious trolling. Therefore, with such values, the test is suitable for assessing peak performance on a short distance.

Results were obtained:

CPU Speed: Events Per Second: 538.23 General Statistics: Total Time: 10.0188S Total Number of Events: 5393 Latency (MS): Min: 19.85 AVG: 22.27 Max: 30.56 95th Percentile: 23.10 SUM: 120125.61 Threads Fairness: Events (AVG / STDDEV): 449.4167 / 4.11 EXECTION TIME (AVG / STDDEV): 10.0105 / 0.01

CPU Speed \u200b\u200bEvents Per Secondmeans the number made in the central operations processor per second - the higher the value, the more productive system.

General Statistics Total Time means the total time to perform operations.

General Statistics Total Number of Events means the total number of events performed.

If the system terminates too fast, you can increase the value, for example, up to two hundred thousand events:

Sysbench CPU --CPU-MAX-PRIME \u003d 200000 RUN

Another way to check trottling and processor performance estimates under a long-term load is setting the execution time, in the example below the time is 300 seconds.

SYSBENCH CPU --THreads \u003d 12 --Time \u003d 300 Run

I have when using options -Time and -CPU-Max-Prime CPU Speed \u200b\u200bEvents Per Second differs in tens of times - apparently or some kind of bug in the program, or the program considers for some other rules.

Benchmark in Phoronix-Test-Suite

Install the program:

Sudo Apt Install Phoronix-Test-Suite

Run - for the first time you will need to take a license agreement, so the program will ask the permission to send anonymous statistics:

PHORONIX-TEST-SUITE LIST-AVAILABLE-SUITES

The previous command is output the list of available benchmarks.

Available sets in the PHORONIX Test Suite V8.0.1 version

Pts / audio-encoding - Audio Encoding System pts / chess - Chess Test Suite Processor pts / compilation - Timed Code Compilation Processor pts / compiler - Compiler Processor pts / compression - Timed File Compression Processor pts / computational - Computational Test Suite System pts / computational -biology - Computational Biology Test Suite Processor pts / cpu - cPU / Processor Suite Processor pts / cryptography - Cryptography Processor pts / daily-kernel-tracker - daily Kernel Tracker system pts / daily-system-tracker - daily system Tracker system pts / database - Database Test Suite System PTS / Desktop-Graphics - Desktop Graphics System PTS / DISK - Disk Test Suite Disk PTS / Encoding - Encoding System PTS / Favorites - Favorites System * PTS / Gaming - Gaming System PTS / Gaming-Closed - Closed-Source Gaming System * PTS / Gaming-Free - Free Software Gaming System PTS / GUI-Toolkits - GUI Toolkits Graphics PTS / IOOQUAKE3-Games - Games using the ioquake3 Engine System PTS / IQC - Image Quality Comparison Suite Graphics PTS / Java - Java System PTS / Java-OpenGL - Java OpenGL System PTS / Kernel - Kernel System PTS / Linux-System - Linux System System PTS / Machine-Learning - Machine Learning System PTS / Memory - Memory Test Suite Memory PTS / Mesa - Mesa Test Suite Graphics PTS / Motherboard - Motherboard System PTS / MultiCore - Multi-Core Processor PTS / Netbook - Netbook Test Suite System PTS / network - Networking Test Suite Network pts / nevada - OpenSolaris nevada Test Suite System pts / opencl - OpenCL System pts / opengl-demos - openGL demos Test Suite System pts / opengl-workstation - openGL Workstation Test Suite System pts / pts-desktop-live - PTS Desktop Live Syste M PTS / Ray-Tracing - Ray-Tracing Test Suite System PTS / Server - Server Motherboard System PTS / UE4 - Unreal Engine 4 Tech Demos On Linux Graphics PTS / Unigine - Unigine Test Suite Graphics * PTS / Universe - Universe Test Suite System PTS / Universe-Cli - Universe CLI Test Suite System * PTS / Universe-X - Universe X Test Suite System PTS / Video-Encoding - Video Encoding System PTS / Workstation - Workstation System PTS / Workstation-Graphics - Workstation Graphics Graphics PTS / Xrender - X Render Extension Testing Graphics System / Collection - Collection System

Star marked partially supported sets.

To start the performance estimate of the central processor, perform:

PHORONIX-TEST-SUITE RUN PTS / CPU

Please note that PTS / CPU and other benchmarks occupy several gigabytes disk space. For example, PTS / CPU will load about 3 GB of data and will use approximately 7 GB of disk space (in the home directory of the user).

About how to control the current frequency and processor temperature in Linux see.

Program cPUBURN. Designed to carry out the stress test of your system by maximum download and warming up the central processor. Optimized for different types processors. It works by performing an infinite cycle for ALU (arithmetic and logical processor unit) and FPU (floating point operations block). Does not test each instruction of the processor, because The goal of the program to warm up the processor as much as possible, thereby giving the stress load on the processor itself, the cooling system, motherboard (especially voltage regulators) and power supplies.

The program consists of the following modules:

burnP5- optimized for Intel processors without MMX instructions;
burnp6. - for Intel processors with support for MMX instructions;
burnk6 - for aMD processors Series k6;
burnk7 - for processors AMD Athlon., Duron and Next ...;
burnmmx. - Test cache / memory interfaces for all processors with support for MMX instructions;
burnBX - Alternative cache / memory test for Intel processors.

Installation:

aPT-Get Install Cpuburn

Fry percent

To load all the kernels - run the desired number of tests of the test suitable for the processor (in my case, the Intel processor has 8 cores):

burnp6 & Burnp6 & Burnp6 & Burnp6 & Burnp6 & Burnp6 & Burnp6 & Burnp6 &

You do not like the nominal limitations of the processor speed, do you prefer to dispersed the processor to the limit of its capabilities? If the processor can work faster, it means that you can try to enlarge for memory clock frequency Tires and / or multiplier. Open processor simply - you need to increase its operating frequency. The acceleration of the frequency of the processor (overclocking) can lead to the unstable operation of the computer, and sometimes, that however, quite rarely happens, it can lead and at all to the output of the computer. Well, since you declined .. To determine the correctness of the processor, in case of overclocking, it will be useful to carry out a test stability of the system at maximum, peak processor employment and the load on RAM. In other words, find out whether the processor errors will appear and the RAM in mode increased productivity Computer. The LINX program is a stress test and for CPU and for RAM.

LINX System Stability Test

Before testing the processor and RAM using resource overload, it is recommended to close all programs to use as much free memory as possible. LINX System Stability Test is implemented using intel technology Linpack. The LINX stress test utility works with any Intel and AMD processors, for Intel processors, the application supports Intel Hyperthreading technology. LINX can test the system stability in 32 and 64-bit test mode.

Settings and LINX capabilities

use to load Intel LINPACK package resources;
support for Intel Hyperthreading technology (Pentium IV, Atom, Core i7);
support for Intel and AMD processors;
processor load in several streams;
choice of priority;
automatic saving settings after the test;
selection of the allocated amount of RAM for test;
preliminary time calculation for stress test;
setting the maximum volume of the LIN32 problem;
error control with the ability to automatically stop stress test in case of error;
automatic saving test report;
alarm notification sound;
import of processor temperature values, CPU fan speed and current voltage values eVEREST programs and Speedfan;
building graphs;
command line support;
creating and maintaining screenshots.