Home

MemTest-86 User Manual Version 6.2

1. Copyright 2015 Passmark Software Page 37 Appendix B Product Support Please report problems to Email helo passmark com Please include The software version A detailed description of the problem and how to reproduce it A copy of the result files log files if available Details of how we may be able to contact you B 1 Known Problems MemTest86 can not diagnose many types of PC failures For example a faulty CPU that causes Windows to crash will most likely just cause MemTest86 to crash in the same way With some PC s MemTest86 will just die with no hints as to what went wrong Without any details it is impossible to fix these failures Fixing these problems will require debugging on your part There is no point in reporting these failures unless you have a Linux system and would be willing to debug the failure MemTest86 supports all types of memory If fact the test has absolutely no knowledge of the memory type nor does it need to This not a problem or bug but is listed here due to the many questions about this issue B 1 1 UEFI v5 MemTest86 depends on the services provided by the UEFI firmware which include mouse keyboard support multiprocessor services file input output and PCI device management Because the services that are provided by the firmware are developed individually by the motherboard manufacturers there may be functional differences than can limit or prevent the proper operation of MemTest86 Some
2. O New CPU 1 a Alddress Range Parallel A11 CPUs i Round Robin All CPUs sequential All CPUs CPU Selection i tart Test RAM Benchmark aS S e ttings le EG it j MemTest8b V6 0 0 Pro Single specifies a single CPU to test and prompts the user to enter a valid CPU number Parallel executes the memory tests on all CPUs concurrently on a set of non overlapping memory segments Round Robin only one CPU is running a test at any given time but cycles to the next CPU in a round robin fashion after every test Sequential only one CPU is running a test at any given time but cycles to the next CPU after a certain memory size has been tested by the CPU Copyright 2015 Passmark Software Page 14 2 4 1 5 RAM Benchmark The RAM Benchmark screen allows the user to benchmark their RAM modules and save the results to disk The benchmark results can then be plotted onto a graph for comparison oy Susten Daf Use the Up 7 Down arrow Keys to select a parameter to change Ag Susten nto a Press lt enter gt or click to change the selected parameter Dest Selection Block Read Write Write Test Mode Memory speed per block size Access Data Width 6 bit TEE Address Range View Prev Results l OPU Selection Available Memory 536412160 Bytes Block Size 2604395456 Bytes S tart Test RYAN Benchmark Test Complete 100 did Sle ttings Press any key to view the results le EG it
3. Dnf Use the Up 7 Down arrow Keys to select a setting to change eo de Pla uystem nio A Maximum system address Ox1FFFOOOO Y T est Selection Enter an address or tEsr to cancel Press lt enter gt or click to change the selected setting U Eg Hex or decimal address 0x10000 4096 Bee Address Range Eg Suffixes k m g 5m 5 megabytes _ LATOR New Value Ox OPU selection Upper Linit Ox AFEOOOO y Reset limits to default gt S tart Test RYAN Benchmark 2 e ttings Il EG it MemTest8b W6 0 0 Pro To change the lower or upper limit of the address range to test select the appropriate setting and enter a new value To enter a decimal address enter the address as is To enter a hexadecimal address enter 0x followed by the hex address You may also use the suffixes k m g to specifiy kilobyte megabytes and gigabytes respectively Selecting Reset Limits to default will reset the limits to its maximum values Copyright 2015 Passmark Software Page 13 2 4 1 4 CPU Selection The CPU Selection screen allows the user to specify a single CPU to test or cycle through all CPUs using a selection method Use the Up Down arrow Keys to select a CPU selection mode a a h 274 ge system TL nfo Press lt enter gt or click to set the CPU selection mode xeWarning Your UEFI firmware has limited multiprocessor support Dest Selection List of CPUs available for testing 0
4. Fixed a bug in test 8 causing a single error to cascade into multiple errors Fixed a bug causing the average error bits to be incorrect once the errors had maxed out at 65k Fixeda bug preventing test 10 to be selected as a single test to run Fixed bug displaying individual test error counts Fixed bug making overall errors 10x what they should be Enhancements in v4 2 Mar 2013 Fixed issues with USB keyboards The USB keyboard functionality is memory mapped into a portion of low memory on some maybe many machines typing on a USB keyboard changes some values in RAM as the key presses are stored in memory as you type This can cause the keyboard to become unresponsive during testing or input from the keyboard to generate errors in the tests Fixed crash when configuring memory ranges Changing the meory range during the first test or changing the memory range multiple times during later tests could cause the current test number to become negative triggering a crash Fixed highest error address not reporting correctly on error Fixed error counters overflowing and reseting to 0 after too many errors Fixed max contiguous error reporting Cleaned up some UI text The Windows USB package now includes ImageUSB to make creating Memtest86 Copyright 2015 Passmark Software Page 48 USB drives easier Enhancements in v4 1 Jan 2013 Added a new boot trace option that single steps through the testing process and displays messages a
5. Limited support for execution with multiple CPUs CPUs are selected round robin or sequential for each test Support for additional chipsets from Memtest86 v2 11 Additions and corrections for CPU detection including reporting of L3 cache Reworked information display for better readability and new information Abbreviated iterations for first pass Enhancements to memory sizing Misc fixes and code cleanup Enhancements in v3 4 8 Sep 2007 A new error Summary display with error confidence analysis Display of memory module information from Memtest86 v1 70 Relocated testing reworked to overlap main testing for better error detection Support for additional chipsets from Memtest86 v1 70 Additions and corrections for CPU identification Misc bug fixes and code cleanup Enhancements in v3 3 12 Jan 2007 Added support for additional chipsets from Memtest86 v1 60 Changed Modulo 20 test 8 to use a more effective random pattern rather than simple ones and zeros Fixed a bug that prevented testing of low memory Added an advanced menu option to display SPD info only for selected chipsets Updated CPU detection for new CPUs and corrected some bugs Reworked online command text for better clarity Added a fix to correct a Badram pattern bug Copyright 2015 Passmark Software Page 51 Appendix D Acknowledgments D 1 UEFI v5 Mem Test86 UEFI was developed by PassMark Software based on the original source code by Chris Br
6. third party to do the actual repair 4 1 Anti Static Handling Procedures Electronic components may be damaged by static electricity The key to proper handling is to simply avoid causing a static discharge though the component that is being handled This is done by discharging static buildup before a component comes in contact with another surface For example when installing a component into a computer first touch a bare metal part of the computer case If you want to set a component on a table touch the table first and then set down the component If you take a step or even shuffle your feet you will need to re discharge any static buildup 4 2 Re Seating Memory Modules In many cases memory problems are caused by a poor connection This can be resolved by simply removing and reinstalling the memory module s using the following procedure 1 Using the documentation for your motherboard locate the memory module slots and identify the memory module s 2 Using proper anti static handing procedures remove the memory module s In most cases this is done by pressing down on the locking tabs at the end of the module slot 3 Carefully clean any dust or debris from the module and the motherboard memory module slots 4 Reinstall the memory module s into their original slot on the motherboard 5 Rerun MemTest86 4 3 Replacing Modules If re seating memory modules does not resolve the problem then a memory module will generally need
7. Hammer Test Choose RAM modules of different brand model as it is likely that the RAM modules with the same model would still fail the Hammer test For sensitive equipment requiring high availability reliability you would replace the RAM without question and would probably switch to RAM with error correction such as ECC RAM Even a 1 bit error can result in catastrophic consequences for say a bank account balance Note that not all motherboards support ECC memory so consult the motherboard specifications before purchasing ECC RAM Copyright 2015 Passmark Software Page 27 4 Repairing Memory Faults When MemtTest86 detects errors the error count will be incremented and the error details will be displayed on the lower half of the screen The key information needed to diagnose errors are the Error Confidence Value and the Errors per Memory Slot information Error confidence values over 100 should always be considered to be legitimate The errors per memory slot indicate both the number of memory modules present in your PC and the number of errors for each memory module This information may not be available for older PCs The remaining error details may be ignored To diagnose and repair a memory fault requires you to open your computer case and handle sensitive electronic components With proper handling procedures this is not difficult If you do not want repair your own hardware you can use Mem Test86 to test your RAM but rely on a
8. LY A V6 0 0 Pro Block Read Write specifies whether memory read or write performance should be benchmarked Test Mode specifies which of the following tests to perform Memory Speed Per Access Step Size This test measures the memory speed with respect to memory accesses of varying distances This test runs through a block of memory sequentially accessing every address Next it runs through the same block again except this time it accesses every second address step size 2 Then every fourth address step size 4 is accessed and so on until a certain maximum step size is reached We should expect to see a decline in memory speed as the distance between memory accesses increase Memory Speed Per Block Size This test measures the memory speed with respect to varying memory block sizes On each subsequent iteration the block size is increased until a certain maximum block size is reached Typically a drop in speed shall be observed when the block no longer fits in Copyright 2015 Passmark Software Page 15 the respective cache levels resulting in the much slower access to main memory Access Data Width Memory Speed Per Block Size only The size of data in bits 8 16 32 or 64 bits to access at one time The best results will usually be obtained when selecting an access size that matches the system s native mode Start performs the benchmark and saves the result to disk under Benchmark Pressing Esc will cancel
9. The log file will be analyzed to determine whether or not the chipset can be supported in future versions View memory usage displays how the system memory address map is allocated amongst the subsystems Memory caching Enabled Disabled Pro version only disable enable memory caching when the tests are running Warning disabling memory caching greatly reduces the performance of the entire system including screen updates Copyright 2015 Passmark Software Page 10 ECC polling Enabled Disabled if ECC detection correction is supported and enabled this option disables enables periodic checking of any ECC errors that have been detected by the system while the memory tests are running ECC injection Enabled Disabled Pro version only if ECC detection correction is supported enabled and ECC injection is supported by the system this option enables disables injection of ECC errors to simulate how the system responds to real ECC errors Note Although ECC injection may be supported by your hardware it may be locked by the BIOS Some BIOS may allow you to unlock the ECC injection feature in the BIOS setup In some cases however you may need a modified BIOS which does not lock the ECC injection feature Save system information summary to file Pro version only Save the system information to a file on disk Copyright 2015 Passmark Software Page 11 2 4 1 2 Test Selection The Test Selection screen allows the user to sel
10. USB Flash drive 1 2 Download the Linux Mac MemTest86 USB image Untar the package tar xvzf memtest86 usb tar gz An image file and a README file will be created in the current directory Follow instructions in the README to write the USB flash disk 2 3 Building MemTest86 v4 BIOS from source Building MemTest86v4 from source requires the GCC compiler that is included in many Linux distributions It should successfully compile for most gcc installs though it has only been tested on gcc 4 7 2 included with Ubuntu 12 10 To compile Mem Test86v4 1 Review the Makefile and adjust options as needed 2 Type make This creates a file named memtest bin which is a bootable image If you encounter build problems a pre compiled binary image has been included precomp bin This image file may be copied to a floppy disk or may be loaded from a disk partition by Lilo or Grub boot loaders from a hard disk partition The Makefile creates bootable images that may be written directly to a floppy disk or CD A method is not provided for creating a bootable USB flash drive from source Copyright 2015 Passmark Software Page 6 Create a MemTest86 floppy disk 1 Insert a blank write enabled floppy disk 2 As root type make install Create a bootable CD ROM 1 type make iso 2 An ISO image file memtest86 iso is created in the source directory Use CD burning software available on your system to create a CD ROM using the ex
11. barriers before and after performing the memory speed test Fixed an error in setting the barrier structure s base address preventing a possible crash or freeze of the system Added a check to perform a spin lock only when more than 1 CPUs are detected Enhancements in v4 3 5 Oct 2013 Fixed potential error due to barrier structure located at fixed memory location Fixed block move test freeze on higher memory addresses Enhancements in v4 3 4 Oct 2013 Fixed incorrect progress calculation for test 0 Fixed incorrect memory size due to bug with memory map when the e820 entry size member is 0 Fixed incorrect number of CPU s found due to duplicate entries in the MADT Changed the method used to search for processors to searching the APIC MADT first then search the MP spec table as opposed to vice versa The MP spec table has largely been deprecated Enhancements in v4 3 3 Sept 2013 Fixed incorrect progress calculation for test 4 Fixed potential false positives in parallel mode caused by overlapped unaligned memory chunk allocations per CPU Fixed program freeze when selecting test 0 or 1 when running in non parallel mode Enhancements in v4 3 2 Aug 2013 Memory bandwidth is now measured for one CPU as opposed to being a total for all CPUs amp Cores This will lower the reported bandwidth for multi core machines But we think it makes more sense this way Fixed crash when attempting to boot on older single core machines with hyp
12. in unreliable operation A much better philosophy is to adjust and fine tune computer timings to maintain an appropriate margin for error An appropriate margin includes not leaving too much margin wasting performance For example let say we have a computer that will operate properly with a system clock of 189 MHz Obviously a lot of performance will be wasted if we operate the computer at the standard 133 MHz In addition if the computer fails at 190 MHz it would be unwise to operate at 189 MHz since there is little margin for error and operation will likely not be reliable An operating margin of 3 to 6 is sufficient to insure good reliability For this example a system clock of 178 to 183 would be ideal There may be cases where it will be advisable to under clock If a system with a normal clock rate of 133 MHz does not operate properly at 136 MHz or more then there is not enough margin and under clocking is required to ensure reliability To summarize over clocking should be thought of as fine tuning a computer to ensure reliability first and secondarily maximizing performance 5 3 Using MemTest86 for Over Clocking The first step in a proper over clocking procedure is to determine the operational limits of your computer After the operational limits have been identified the system settings may be adjusted to provide enough margin to ensure high reliability MemTest86 is an ideal tool to accurately determine the operation limits of your memor
13. of these issues are highlighted below Some UEFI implementation do not provide full mouse support preventing the use of the mouse in MemTest86 If this is the case use the keyboard or try updating to a new firmware build Multiple CPU testing may not be available due to limited or non functional implementations of Multiprocessor services provided by UEFI especially for older firmware This may cause a reduced number of processors available for testing or even program freeze when attempting to run on other processors If this is the case run in single CPU mode or try updating to a new firmware build MemtTest86 is also inherently limited by the UEFI environment and as such gives rise to the following limitations MemTest86 USB flash drives cannot be read in Windows XP 32 bit due to its lack of support of GP Tdisks MemTest86 cannot remap itself to different portions of memory in order to run tests in Copyright 2015 Passmark Software Page 38 the section of memory it was occupying Dual UEFI entries may be present in UEFI BIOS as boot devices There is no difference in selecting either entry This is due to a workaround that allows MemTest86 USB flash drives to be accessible in Windows B 1 2 BIOS v4 sometimes when booting from a floppy disk the following messages scroll up on the screen X 8000 AX 0212 BX 8600 CX 0201 DX 0000 This is the BIOS reporting floppy disk read errors Either re write or toss the floppy disk On a
14. or reduced iteration New config file parameter ADDR2CHBITS to specify the address bits to XOR to determine the memory channel New config file parameter LANG for specifying language to use on startup Console resolution is now forced to 80 x 25 Graphics resolution is now set to a minimum of 1024 x 768 Updated ImageUSB to v1 1 1015 which includes an option to zero the USB drive Running memory tests in parallel mode is now more robust for UEFI BIOS that exhibit inconsistent multiprocessor behaviour Fixed detection of the number of enabled processors for UEFI BIOS that exhibit inconsistent multiprocessor behaviour Fixed test status screen not being displayed correctly for consoles with small large screen widths In the RAM SPD menu screen PGUP PGDN can be used to navigate between pages of RAM modules For specific cases where files under EFI BOOT cannot be accessed eg grub2 log report files shall be written to the root directory During MemTest86 boot up the system memory map is now written to log file Various optimizations of the size of the MemTest86 binary Forced a memory address limit of 32 bits when running under 32 bit UEFI Memory ranges to be tested are now allocated at the beginning of each test due to the possibility that the memory map changes in the middle of testing Reduced the number of log messages written when waiting for other processors to Copyright 2015 Passmark Software Page 43 finish when running in
15. small number of machines false positives were reported when running Test 3 in parallel mode If this is the case re run the test in single CPU mode before concluding that it is a RAM issue B 2 Enhancements Please send enhancement requests to info passmark com All requests will be considered but not all can be implemented Copyright 2015 Passmark Software Page 39 Appendix C Change Log Changes in v6 2 0 Sept 2015 New Features Due to the high number of failures reported for the Hammer Test Test 13 the algorithm was revised to perform 2 potential passes 1 Row pairs are hammered at the maximum hammer rate ie no delays between each row pair hammer 2 Row pairs are hammered at a lower hammer rate 200K per 64ms as determined by memory vendors as the worst case scenario o If memory errors are detected in the first pass error details are not immediately displayed to the user and the second pass is started If errors are detected in the second pass they are reported as normal o f errors are detected in the first pass but not the second pass a warning of potential high frequency bit flips is displayed to the user o The premise behind this revision is to better inform users of the significance of errors detected in the Hammer Test as opposed to a strict PASS FAIL result Although errors detected in this test are real errors the conditions needed to induce these errors occur only very rarely in normal PC usage and
16. system clock rate by 3 to 6 from the maximum setting that functioned properly In many cases the operational limits for the CPU and memory will be different For example you can get more CPU speed by reducing memory settings Or you can get more memory performance by reducing the CPU multiplier For these cases you will need to choose a compromise Both memory bandwidth and CPU clock rate are important so don t be tempted to only optimize for one or the other Mem Test86 provides good assurance of reliable memory operation when over clocking Even when a dramatic increase in memory bandwidth is achieved As long as Mem Test86 does not report errors and appropriate margins have been applied then you should not hesitate to fully maximize memory timings However some caution must be exercised for system clock rate increases With most motherboards the clock rate for the PCI and AGP busses are based on the system clock Generally these busses will have no problem running at somewhat higher rates MemTest86 does not test PCI or AGP and do not provide any assurance that anything other than the CPU and memory are working properly Sadly there is currently no safe way to determine the operational limits for PCI and AGP and therefor there is no way to assure that there are appropriate margins Unless your motherboard is able to independently establish the frequency of PCI and AGP busses you should be careful about running with large more than 15 increases
17. that all of the tests have been run The first pass is abbreviated to find errors more quickly Subsequent passes execute longer and will be more thorough Generally a single pass is sufficient to detect all but the most obscure errors For Copyright 2015 Passmark Software Page 22 complete confidence in cases where intermittent errors are suspected testing for a longer period is advised The time required for a complete pass of MemTest86 will vary greatly depending on CPU speed memory speed and memory size If memory errors are detected they will be displayed on the lower half of the screen The default error reporting mode will display a detailed summary of all errors If MemTest86 runs multiple passes without errors you can be certain that your memory is functioning properly Other problems hardware or software may exist but at least you can eliminate memory as a Culprit In addition the CPU must work properly to run MemTest86 so successful execution implicitly assures that your CPU is also functioning properly MemTest86 can not diagnose many types of PC failures For example a faulty CPU that causes Windows to crash will most likely just cause Mem Test86 to crash in the same way 2 5 1 1 Runtime Configuration Options MemTest86 may be configured during operation via runtime configuration commands Pressing the C key at anytime will display the runtime command menu Tthe following options are available in the Configuration command men
18. the test in the middle of testing View Prev Results allows the user to select from a list of benchmark results saved under Benchmark to plot on a graph ID Susten Dnf Use the Up 7 Down arrow Keys to move between the results ALS system nio Press lt enter gt or click to include exclude the result from the graph me Address Range O PU Selection Dest Selection S tart Test MAM Benchmark As e ttings 2014 11 18 2014 11 13 2014 11 13 2014 11 13 2014 11 13 2014 11 13 2014 11 14 2014 11 17 2014 11 19 014 11 19 2014 11 13 2014 11 19 2014 12 12 2014 12 12 2914 12 12 2014 12 15 Read F Read Write Write Write Write Write yRead Read Read Read Write Write Write Write Read b4bit 511Mb b4bit 511Mb 16bit Z 511MB 16bit Z 511MB b4bit 511Mb 16bit Z 511MB b4bit 511Mb b4bit 511Mb 16bit Z 511MB 16bit Z 511MB Obit Z 511Mb Obit Z 511Mb b4bit Z SLIME 64bit Z 511NB 32bit Z 511Mb b4bit 511Mb le E bd it amp MenTest86 V6 0 0 Pro Use the up down arrow keys or mouse to highlight a result to graph then press enter or click Amaximum of 8 results can be plotted onto a single graph Once all desired results are selected selecting Graph selected result s shall plot the selected result s to a graph Copyright 2015 Passmark Software Page 16 The graph can be saved to a bitmap file or exporte
19. to use some patterns to insure that all adjacent cells have at least been written with all possible one and zero combinations It can also be seen that caching buffering and out of order execution will interfere with the moving inversions algorithm and make less effective It is possible to turn off cache but the memory buffering in new high performance chips can not be disabled To address this limitation a new algorithm called Modulo X was created This algorithm is not affected by cache or buffering The algorithm works as follows 1 For starting offsets of O 20 do steps 2 5 2 write every 20th location with a pattern 3 write all other locations with the patterns complement 4 repeat 1b one or more times 5 check every 20th location for the pattern This algorithm accomplishes nearly the same level of adjacency testing as moving inversions but is not affected by caching or buffering Since separate write passes 1a 1b and the read pass 1c are done for all of memory we can be assured that all of the buffers and cache have been flushed between passes The selection of 20 as the stride size was somewhat arbitrary Larger strides may be more effective but would take longer to execute The choice of 20 seemed to be a reasonable compromise between speed and thoroughness A 3 Individual Test Descriptions MemTest86 executes a series of numbered test sections to check for errors These test sections consist of a combination of test algorith
20. 3 albeit exposed only in extreme memory access cases are most certainly real errors During typical home PC usage eg web browsing word processing etc it is less likely that the memory usage pattern will fall into the extreme case that make it vulnerable to disturbance errors It may be of greater concern if you were running highly sensitive equipment such as medical equipment aircraft control systems or bank database servers It is impossible to predict with any accuracy if these errors will occur in real life applications One would need to do a major scientific study of 1000 of computers and their usage patterns then do a forensic analysis of each application to study how it makes use of the RAM while it executes To date we have only seen 1 bit errors as a result of running the Hammer Test There are several actions that can be taken when you discover that your RAM modules are vulnerable to disturbance errors Do nothing Replace the RAM modules Use RAM modules with error checking capabilities eg ECC Depending on your willingness to live with the possibility of these errors manifesting itself as real problems you may choose to do nothing and accept the risk For home use you may be willing to live with the errors In our experience we have several machines that have been stable for home office use despite experiencing errors in the Hammer Test You may also choose to replace the RAM with m odules that have been known to pass the
21. LT Option key on the Mac keyboard while powering on the machine All MemTest86 images supports dual booting of MemTest86 v4 BIOS and MemTest86 v5 or later UEFI depending on whether your system is configured to boot in UEFI or BIOS mode If your system is booting MemTest86 v4 it is most likely that either 1 You have an older system that does not support UEFI 2 Your system supports UEFI but is configured in legacy mode ie BIOS If 1 is true you system will not be able to boot MemTest86 v5 or later You will need to upgrade to a new system that supports UEFI in order to run MemTest86 v5 or later If 2 is true you will need to go to the BIOS setup and change the necessary settings in order to boot from UEFI The actual BIOS setting varies depending on the vendor but it is typically Legacy Boot CSM or Compatibility Support Module Copyright 2015 Passmark Software Page 8 When Mem Test86 boots a splashscreen is displayed with a 10 second countdown timer which when expires automatically starts the memory tests with default settings Pressing a key or moving the mouse shall stop the timer Memory tests will automatically start in 1 2 seconds Please select an icon to continue Use the Left Right arrow keys or mouse p Config Selecting Exit shall reboot your system To configure the memory tests select Config and the main menu is displayed The main menu allows the user to customize the memory test setting
22. MemTest 86 User Manual Version 6 2 Copyright 2015 Passmark Software Page 1 Table of Contents PERO GUC r ela RONA adie N a R 3 B Ory REM N NN PN DD dd le ad ad le 3 12 MenilestS K lo AA E EEE dee a pa j a a 3 bte e br rJJJJ rr r r rmrm _eaUr m m_m__ooB xy a 3 ZU ad BD Sc oo ee wvvwvwvwve9 5 2 1 Boot disk Creation using WindOWS keke keke Ke Keke kek kek kK KK KAK KK KK KA 5 2 2 DOOt5QISK Creation usine Lina Macia E 6 2 3 Building MemTest86 v4 BIOS from SOUTCe ELL kk K k K KK KA 6 24 BS ne Vol Les TO UEF a Ae Y 8 2 5 sine Mem LESS A BO e cues ahi 22 3 Troubleshootina Memory ETS id ld ias 26 i Hammer Test esto PROS Ai 26 4 Repani ne Memory Ea aaa 28 4 Amita static Handlmo Proceduro S tddi 28 42 Re gt cata Memory Modules aa speeds eet Key a Sedan de 28 A REP LAC ne MOM Sia ai 28 FAE ror NAG N Ze r bacco ag hd woolen dae ag ses eee 29 FOVET CIOCKING NNN tack de tal ct Dantas delat cit aK de tal ot r F a aR ra re 30 PR u ooo en ae ee 30 2 OPS MATAS ao 30 5 Using MemTestso tor Over Clockin ereinen enn Ne AE E anime 30 ea 0151016 A O O A ge ree 33 Appendix A lechnical torn 33 As Memory resto PililoS pDl y y yc i Ar ye ya k seyan E deis 33 A2 MEMES LESA OS 33 A Individual Test Description aaa 34 A4 Error Display MOMO nia ide 36 Appendix B Product SUD DOM ca ce eae 38 BAR TONE PODIO ND D _ _eb bby j rWm mw mwmW
23. NG ja JP REPORTNUMERRS 10 REPORTNUMWARN 10 AUTOMODE 1 Lines that start with indicate a comment line All parameters are specified as follows Parameter name Parameter value The following table summarizes the list of supported parameters Parameter Description TSTLIST List of tests to execute in the test sequence Each test is specified by a test number separated by a comma NUMPASS Number of iterations of the test sequence to execute This must be a number greater than 0 ADDRLIMLO The lower limit of the address range to test To specify a hex address the address must begin with Ox Otherwise the address shall be interpreted as a decimal address ADDRLIMHI The upper limit of the address range to test To specify a hex address the address must begin with Ox Otherwise the address shall be interpreted as a decimal address CPUSEL One of the following CPU selection modes SINGLE PARALLEL RROBIN SEQ CPUNUM The CPU of the specific CPU to test This parameter only has an effect if CPUSEL is set to SINGLE ECCPOLL Specifies whether ECC errors shall be polled 0 Polling disabled 1 Polling enabled ECCINJECT Specifies whether ECC error injection shall be enabled 0 ECC injection disabled 1 ECC injection enabled Copyright 2015 Passmark Software Page 18 MEMCACHE Specifies whether memory caching shall be enabled disabled during te
24. a i r rn 38 BZ ETI Ran ce MCRL ut did 39 Pope Nai C MDEE OO caress SAA A A 40 Appendix DA Chow led omens adictas 50 POT RET VO un tias 50 DABIOS Va a caleba O 50 Copyright 2015 Passmark Software Page 2 1 Introduction 1 1 Memory Reliability Properly functioning memory is critical for reliable operation of a PC or laptop Few computer users fully understand the risks associated with memory errors Because PCs typically do not have any mechanisms for detecting memory errors confusing and potentially disastrous consequences can result from these undetected memory problems Memory errors will often cause erratic behavior with software applications that can mysteriously fail The most serious risk from memory errors however is corruption of data that manages how information is stored on disk In most cases this type of corruption will cause one or more files to be lost There are cases where a memory error can cause the loss of the entire contents of your hard disk Periodic testing of memory with a rigorous and thorough memory test will greatly reduce the risk of problems and data loss due to memory errors 1 2 MemTest86 Overview Memory errors are often pattern sensitive and may be very intermittent Detecting these errors is technically challenging and is an imperfect science MemTest86 uses advanced algorithms that have been refined for more than 20 years These testing techniques are highly effective at detecting difficult to fin
25. ady Hammer Test was implemented based on the research work performed and documented in Flipping Bits in Memory Without Accessing Them An Experimental Study of DRAM Disturbance Errors by Yoongu Kim et al Translation work performed by Freelancer members GauthierC French daufenbach German Kentaro Ide Japanese tomhu Chinese Spanish translations provided by Gabriel Barrandeguy of GabakTech D 2 BIOS v4 MemTest86 was developed by Chris Brady and maintained by PassMark Software with the resources and generous assistance of individuals and sources listed below The initial versions of the source files bootsect S setup S head S and build c are from the Linux 1 2 1 kernel and have been heavily modified Doug Sisk provided code to support a console connected via a serial port Code to create BadRAM patterns was provided by Rick van Rein Test 5 is based on Robert Redelmeier s burnBX test Screen buffer code was provided by Jani Averbach Eric Biederman provided all of the feature content for version 3 0 plus many bugfixes and significant code cleanup Major enhancements to hardware detection and reporting in version 3 2 3 3 and 3 4 provided by Samuel Demeulemeester from MemTest86 v1 11 v1 60 and v1 70 Copyright 2015 Passmark Software Page 52
26. anufacturer ID lookup function to support gt 7 continuation codes Fixed AMD Hudson 2 Hudson 3 SMBus support to include various hardware revisions Various system info related updates fixes SPD CPU Changes in v6 1 0 June 2015 New Features New config file parameter REPORTNUMERRS for specifying the maximum number of errors to include in the report Language support for Spanish RAM details obtained from SMBIOS are now displayed if SPD information cannot be accessed Fixes Enhancements Added ECC detection support for Intel Atom C2000 SoC chipset Added ECC injection support for Intel Atom C2000 SoC chipset Added ECC injection support for Intel Xeon E3 Sandy Bridge lvy Bridge chipset untested Added SMBus SPD support for Intel 5100 chipset Added SMBus SPD support for Intel Wildcat Point chipset Added SMBus SPD support for Intel Sunrise Point chipset Improved speed of retrieving SPD data Fixed potential issue with retrieving DDR4 SPD due to the bank address not being restored back to its original value Fixed decoding of ECC support in DDR2 SPD Added support for more precise timings supported by DDR3 rev 1 1 and later Fixed bug with retrieving SPD details for a large number of RAM modules ie gt 16 Fixed the progress indicator for Test 13 Row hammer test to be more linear Reduced the test time for Test 13 Row hammer test Fixed the displayed error details when ECC errors are detected Fixed freeze while initiali
27. be specified separated by spaces For example memtest onepass tstlist 1 6 7 8 will start the test with the One Pass option enabled and execute tests 1 6 and 8 2 5 3 Troubleshooting with Boot Trace This is section is targeted at troubleshooting problems with execution of the test code and not for diagnosing reported memory errors In Version 4 1 a Boot Trace feature was added that provides a simple mechanism for troubleshooting of test failures by both technical and non technical users With the very large variety of computer hardware available it is impossible to test Memtest86 an all platforms and some incompatibilities exist causing failures In the past determining the cause of these failures has been difficult to impossible With trace information many of these faults may now be addressed Enabling Boot Trace The boot trace feature is enabled with either a boot time option or an on line command However most problems occur during startup so the boot option is the preferred method All of the released images have a boot option to start the test with boot tracing enabled When booting from a Linux system with LILO or GRUB add btrace to the boot options line When Boot Trace is enabled two columns of trace information will be displayed one the bottom two thirds of the screen The CPU line number from the source file a short message and two parameters are displayed in each trace point A gt character denotes the current
28. cking can only be done in the forward direction On the first pass a fixed seed number is used so that the random number generator always generates the same sequence of numbers allowing the test run to be reproducible All subsequent passes use a seed number generated from the system clock resulting in more permutations being tested The 64 bit and 128 bit versions of this test is essentially the same as the original 32 bit test except that it uses native 64 bit and SIMD instructions respectively Modulo 20 random pattern Using the Modulo X algorithm should uncover errors that are not detected by moving inversions due to cache and buffering interference with the algorithm A sequence of 6 32 bit random patterns are used Bit fade test 2 patterns The bit fade test initializes all of memory with a pattern and then sleeps for 5 minutes Then memory is examined to see if any memory bits have changed All ones and all zero patterns are used Copyright 2015 Passmark Software Page 35 Row Hammer Test The row hammer test exposes a fundamental defect with RAM modules 2010 or later This defect can lead to disturbance errors when repeatedly accessing addresses in the same memory bank but different rows in a short period of time The repeated opening closing of rows causes Charge leakage in adjacent rows potentially causing bits to flip This test hammers rows by alternatively reading two addresses in a repeated fashion then verifying t
29. cluding Native 64 bit support No longer requires the use of the PAE workaround to access more than 4GB of memory PAE Physical Address Extension Mouse support where supported by the underlying UEFI system On older systems a keyboard is still required Improved USB keyboard support The keyboard now works on systems that fail to emulate lO Port 64 60 correctly So Mac USB keyboards are now supported Improved multi threading support where supported by the underlying UEFI system Reporting of detailed RAM SPD information Timings clock speeds vendor names and much more Support to writing to the USB drive that MemTest86 is running from for logging and report generation In all prior MemTest86 releases there was no disk support Use of GPT GUID Partition Table If UEFI is not supported on the system the older v4 BIOS version is booted Mem Test86 can boot from a CD USB flash drive or with Linux systems by the boot loader for example LILO or Grub Any Windows Linux or Mac system may be used to create the CD or USB flash drive Once a MemTest86 boot disk has been created it may be used on any x86 PC computer with either a CD ROM drive or a USB flash drive 2 1 Boot disk Creation using Windows Create a boot able CD ROM 1 2 Download the Windows MemTest86 ISO image Right click on the downloaded file and select the Extract to Here option This places the CD ROM ISO image in the current folder Use
30. creased the maximum number of SMBus controllers to 8 DDR3 revision 1 3 SPD decoding now supported Fixed SMBUS CLK issues when retrieving SPD details for Intel chipsets CPU spec updates AMD Kaveri Intel Haswell refresh Xeon Ivy Bridge and later non Turbo CPU speeds now recorded Minor temperature reporting changes for AMD Family 15h Models Oh OFh and 30h Copyright 2015 Passmark Software Page 44 3Fh e g A10 7850K Fixed Pass progress bar so that it shows 100 on completion of one pass A notification text is now displayed when ECC errors are injected Improved notificaion text displayed when ECC errors are detected Updated MemTest86 BIOS version to 4 3 7 Copyright 2015 Passmark Software Page 45 New features in v5 0 0 Dec 2013 Completely re written to work under UEFI Native 64 bit support No longer requires the use of the PAE workaround to access more than 4GB of memory PAE Physical Address Extension Mouse support where supported by the underlying UEFI system On older systems a keyboard is still required Improved USB keyboard support The keyboard now works on systems that fail to emulate lO Port 64 60 correctly So Mac USB keyboards are now supported Improved multi threading support where supported by the underlying UEFI system Dual boot with Memtest version 4 for supporting older systems without UEFI So with a single USB or CD drive both UEFI systems and BIOS systems can be supported Reporting
31. d into an HTML file as a report Pro version only The report contains the system information the graph itself and the corresponding raw data eA Systen D nfo Henory speed MNB Sec per Block Size TEN E j Lo eee Read 16bit 12600 A A 511MB A O No NS Read Shit Y Dest Selection HO p TT mj x p EB 511MB T N Write 32bit 511MB E 8400 LEE a TT Read 64bit BIB Address Range 7000 BEM NEN SEI MERI DEM KN MEW EMRE j a 5600 yan 4200 TT TT TT EAGE CPU Selection A A J eS 2800 az 5 NOT O rT 1 EEE En gt 8 tart Test ee 16MB 64MB 256MB Block Size RYAN Benchmark s e ttings Sort results as HTML J E tx it Go back MenTest66 W6 0 0 Pro Copyright 2015 Passmark Software Page 17 2 4 2 Configuration File Pro version only Memory test parameters can also be set via a configuration file mt86 cfg that is loaded on startup without the need to manually configure the memory tests every time MemTest86 is run This is useful especially in testing environments where memory tests need to be executed in an automated fashion without user intervention A sample configuration file is as follows MemTest86 configuration file it ToTETSTSOy 1 a NUMPASS 3 ADDRLIMLO 0x10000000 ADDRLIMHI 0x20000000 CPUSEL PARALLEL CPUNUM 1 ECCPOLL 0 ECCINJECT 0 MEMCACHE 0 PASS1FULL 0 ADDR2CHBITS 12 9 7 LA
32. d memory errors In addition MemTest86 has the capability to test all available memory Memory testing programs execute from memory and therefore are not able to test the memory that is occupied by the test program itself When running the BIOS version Mem Test86 is able to move itself to a different portion of memory and then tests the memory that it previously occupied The UEFI version due to platform limitations is unable to remap itself to different portions of memory in order to run tests in the section of memory it was occupying The UEFI firmware itself also takes up some space compared to a traditional BIOS So slightly less RAM can be tested compared to the BIOS version 1 3 Compatibility MemTest86 is designed to work with all processors using the Intel AMD x86 and X86_64 architecture running on UEFI or BIOS systems Most newer systems are able to run the UEFI version of MemTestg86 but all systems should be able to boot the traditional BIOS version MemTestg86 is able to test all types of memory There is no need for MemTest86 to know what type of memory it is testing MemTest86 attempts to detect and display information about the hardware it is testing but this information is not used during testing Since MemTest86 is a standalone program it does not require any operating system support for execution It can be used with any PC regardless of what operating system if any is installed MemTest86 is multi threaded and is able t
33. de that will be used as a header to the test report This can contain a company logo contact information or any additional information about the test environment mt86foot htm The HTML code that will be used as a footer to the test report Examples of usage include a signature line to indicate the technician who performed the test or a disclaimer about the validity of the test results report css The stylesheet used to specify the appearance of the report The file itself contains all the properties that are used along with several templates that can be used to customize the report 2 4 5 Troubleshooting MemTest86 Problems A log file MemTest86 log is automatically created and updated while MemTest86 is running This log file contains information that is helpful in diagnosing possible memory failures or problems with MemTest86 itself If you believe you may have encountered a bug with MemTestg86 please report problems to helo passmark com Copyright 2015 Passmark Software Page 21 2 5 Using MemTest86 v4 BIOS To start MemTest86 insert the MemTest86 floppy disk CD ROM or USB flash drive into the appropriate drive and restart your computer Note The BIOS must be configured to boot from the device that MemTest86 is installed on Newer computers have an optional boot menu that is enabled be pressing a key at startup often ESC F9 F11 or F12 If available use the boot menu to select the correct drive With older computers t
34. dity There is a high probability that all errors reported are valid regardless of this value However when this value exceeds 100 it is nearly impossible that the reported errors will be invalid Lowest Error Address The lowest address that where an error has been reported Highest Error Address The highest address that where an error has been reported Bits in Error Mask A mask of all bits that have been in error hexadecimal Bits in Error Total bit in error for all error instances and the min max and average bit in error of each individual occurrence Max Contiguous Errors The maximum of contiguous addresses with errors Copyright 2015 Passmark Software Page 36 ECC Correctable Uncorrectable Errors The number of errors that have been corrected uncorrected by ECC hardware Test Errors On the right hand side of the screen the number of errors for each test are displayed When the individual error reporting mode is selected the following information is displayed An error message is only displayed for errors with a different address or failing bit pattern All displayed values are in hexadecimal Label Description Tst Test number Failing Address Failing memory address Good Expected data pattern Bad Failing data pattern Err Bits Exclusive or of good and bad data this shows the position of the failing bit s Count Number of consecutive errors with the same address and failing bits
35. ect the test sequence to run and the number of passes to run each sequence See Individual Test Descriptions for a detailed description of each test Re A KA system t I nf 2 Dest Selection f ddress Range OP Selection SS tart Test RAM Benchmark 3 ale tings MemTest8b V6 0 0 Pro Test Test Test Test Test Test Test Test Test Test Test 1 2 3 4 J b T n 3 Use the Up Z Down arrow Keys to move between the tests Press lt enter gt or click to enable disable the selected test g on Address test own address 1 CPU Address test own address Moving inversions ones amp zeroes Moving inversions 6 bit pattern Moving inversions random pattern Block move 64 byte blocks Moving inversions 32 bit pattern Random number sequencel Modulo 20 ones zeros 10 Bit fade test 2 patterns 1 CPU 11 Random number sequence 64 bit 12 Random number sequence 120 bit 13 Hammer test Nunber of passes To enable disable a test use the up down arrow keys or mouse to highlight a test then press enter or click Test 11 and 12 are available only in the Pro version To change the number of passes select Number of passes and enter the desired number of passes Copyright 2015 Passmark Software Page 12 2 4 1 3 Address Range The Address Range screen allows the user to specify a subset of the total system address map to test PD Susten
36. ell E DDR4 ECC support Added Xeon E5 v3 ECC support Added lvy Bridge non Xeon ECC support Added AMD Steppe Eagle ECC support Added Intel Atom E3800 SoC ECC support Fixed ECC detection for Ivy Bridge EX Haswell EX chipsets that have a 2nd memory controller Fixed Intel5400 ECC registers not being reset after starting test Fixed ECC errors immediately being reported after starting test Ivy Bridge E Added support for ECC injection for Intel Xeon E3 v3 untested Fixed handling of Intel ICH SMBUS built in hardware semaphore to prevent SMBus device contention Fixed possible crash when DDR3 module type value in the RAM SPD info is invalid Fixed DDR4 SPD clock speed rounding errors in the RAM SPD info Copyright 2015 Passmark Software Page 42 Fixed DDR3 SPD Register manufacturer type in the RAM SPD info not appearing correctly CPU speed measurement is now more robust by taking multiple samples Fixed Intel turbo clock speed calculation Fixed detection of Intel turbo support for Xeon chipsets Increased maximum of supported CPUs to 72 Increased maximum of supported RAM modules to 64 Increased the number of supported memory controllers to 8 New config file parameter ECCINJECT for specifying whether to enable disable ECC injection New config file parameter MEMCACHE for specifying whether to enable disable memory caching New config file parameter PASS1FULL for specifying whether the first pass should run the full iteration
37. erthreading Only effects old machines from around the early Pentium 4 era that didn t have a MP Multi Processor Spec table defined but did have both a MADT Multiple APIC Description Table defined and hyperthreading enabled Restored the Start only one CPU boot option This option should not be required in normal use but might be useful for debugging purposes Updates to the included help file Copyright 2015 Passmark Software Page 47 Enhancements in v4 3 1 Aug 2013 Fixed bug with Test 6 Block Move Test not testing the end of a memory segment correctly Removed unnecessary boot options in menu Enhancements in v4 3 Jul 2013 Changed default CPU selection mode to round robin Running all CPUs at once has been shown to cause false positives on a number of systems Fixed a bug that could cause the program to go into a tight loop that could not be escaped when setting certain memory ranges to test Fixed a bug displaying the memory location of individual errors The values after the decimal point in the MB readout were incorrect Fixed a bug in configuring upper and lower memory limits previously lower limits equal or grater than 2gb would not work as well as some other more obsucre configurations Added a misc option to display the systems memory map Fixed a bug that would cause the number of passes to not correctly reset after changing the selected tests Added missing source code to some of the download packages
38. h default settings 7 Miscellaneous Options Enable and disable runtime options Currently only two options are available One Pass and Boot trace The One Pass feature runs the complete test once and then exits but only if there were no errors This provides a convenient method for unattended testing The Boot Trace option is a debugging tool that is primarily designed for troubleshooting startup problems but may be enabled at anytime A help bar is displayed at the bottom of the screen with the following options Keyboard Description Assignment ESC Exits the test and does a warm restart via the BIOS C Enter the configuration menu SP Spacebar Set scroll lock Stops scrolling of error messages Note Memory testing is suspended when the scroll lock option is set and the scrolling display region is full CR Enter Clear scroll lock Enables error message scrolling 2 5 2 Startup boot Options A number of options may be specified at boot time The following options are available One Pass Enables the One Pass option This feature runs the complete test once and then exits but only if there were no errors This provides a convenient method for unattended testing One Pass may also be enabled via an on line command The syntax is onepass and there are no additional options eBtrace Enables the boot trace option and is used to diagnose test failures please see section 2 7 for details T
39. he boot sequence must list the drive used to load Memtest86 before any other bootable media i e a hard disk where Windows is installed Most systems will be configured with a suitable boot sequence If not you can reconfigure the boot sequence using the BIOS settings Please consult your motherboard documentation for details 2 5 1 Testing When MemTestg86 starts it displays details about the system configuration and then begins testing MemTest86 executes a repeating cycle of tests Testing will continue to run until the program execution is interrupted by pressing the ESC key or pressing the reset button There is no set time for how long the test should be run The following is an example of the MemTest86 screen AMD FxX tm 41060 Quad Core Processor 4240 MHz Pass 9z HHH Li Cache 00k 34361 MB s Testi Z HHHH HHHH HHH H HHH HE HE HE H HE HE HEHHE HH HHH HHH HHH HHH H L2 Cache 2046K 11971 MBs Test 2 Address test own address Parallel L3 Cache 68192K 4336 MB s Testing Z048M 3584M 1536M of 3730M 9046 MB s Pattern 01234567 CPUs Found CPU Mask 2 8 Lu L y a U a CPUs Started 8 CPUs Active AdrsMode 64Bit Pass ESClexit Ccdconfiguration SP scroll_lock CCRIJscroll_unlock MemTest86 executes a series of numbered test sections to check for errors The execution order for these tests has been arranged so that errors will be detected as rapidly as possible The pass counter increments each time
40. he contents of other addresses for disturbance errors For more details on DRAM disturbance errors see Flipping Bits in Memory Without Accessing Them An Experimental Study of DRAM Disturbance Errors by Yoongu Kim et al Starting from MemTest86 v6 2 potentially two passes of row hammer testing are performed On the first pass address pairs are hammered at the highest possible rate If errors are detected on the first pass errors are not immediately reported and a second pass is started In this pass address pairs are hammered at a lower rate deemed as the worst case scenario by memory vendors 200K accesses per 64ms If errors are also detected in this pass the errors are reported to the user as normal However if only the first pass produces an error a warning message is instead displayed to the user A 4 Error Display Information MemTest86 has 4 options for reporting errors The default is error summary mode where key failure information is displayed along with an error confidence value Reporting of individual errors is also available In BadRAM Patterns mode patterns are created for use with the Linux BadRAM feature This feature allows Linux to avoid bad memory pages Details about the BadRAM feature can be found at PELs home zonnet nl vanrein badram The error summary mode reports the following data Error Confidence Value A value that indicates the validity of the errors being reported with larger values indicating greater vali
41. he syntax is btrace and there are no additional options Maxcpus Set the maximum number of CPU s to start This is primarily a troubleshooting option Setting maxcpus to one skips all code related to finding and starting additional CPUs The syntax is maxcpus N where N 1 to 31 lest List Allows for execution on a subset of the normal test sequence The syntax is tstlist list where list is a comma separated list of test numbers to run do not include spaces Example tstlist 1 4 5 7 CPU Mask A mask of CPU numbers that are started and enables A 32 bit mask is specified with a 1 set for each CPU that will be enabled CPU 0 cannot be disabled and will be enabled regardless of the mask The syntax is cpumask N where N is a 32 bit hexadecimal number with or without a Ox prefix Example coumask 0x55555555 enables all even numbered CPUs Copyright 2015 Passmark Software Page 24 Console Used to setup serial console parameters The syntax is console ttySn baudPL where n serial port number 0 or 1 baud baud rate P parity setting N O or E and Lis the number of bits 7 or 8 Example console ttySO 9600e8 uses port O at 9600 baud even parity and 8 bits The standard Memtest86 media images provided use Syslinux for booting and allow for boot options to be specified interactively At the boot prompt type memtest followed by any of the above options Multiple options may
42. ifies whether to automatically save the test results to the report file and exit MemTest86 ie reboot when the tests have completed O Auto mode disabled prompt user after test completion 1 Auto mode enabled do not prompt user after test completion Copyright 2015 Passmark Software Page 19 2 4 3 Testing Once the memory test has started the following screen which shows the test status is displayed Intel Core i 7 2600 E 3 406Hz Clk Temp 3254 MHz Pass 204 HHHHHHHHE Li Cache 648 79300 MB s Test 96 HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH L2 Cache 6144K 27506 MB s L3 Cache N A Address Ox1D6E5000 0x1DD83000 Memory 511M 12200 MB s Pattern OxEFEFEFEF RAM Info CPUs Found State A CPUs Started 1 CPUs Active Tine 0 00 10 AddrMode 32 bit Pass 1 4 Errors ESC Z c onfiguration MemTest86 executes a series of numbered test sections to check for errors The execution order for these tests has been arranged so that errors will be detected as rapidly as possible The time required for a complete pass of MemTest86 will vary greatly depending on CPU speed memory speed and memory size If memory errors are detected they will be displayed on the lower half of the screen If MemTest86 runs multiple passes without errors you can be certain that your memory is functioning properly Other problems hardware or software may exist but at least you can eliminate memory as a Culprit I
43. in the system clock In addition running your CPU at higher frequencies will generate more heat Small frequency increases will generally be fine with the installed CPU cooler Larger increases in the system clock rate may necessitate a larger more effective CPU cooler Copyright 2015 Passmark Software Page 32 Appendices Appendix A Technical Information Appendix A contains technical information from the MemTest86 README file that was previously released under the Gnu Public License GPL This section provides additional background information and technical information that may be useful for advanced users A 1 Memory Testing Philosophy There are many approaches for testing memory However many tests simply throw some patterns at memory without much thought or knowledge of memory architecture or how errors can best be detected This works fine for hard memory failures but does little to find intermittent errors BIOS based memory tests are useless for finding intermittent memory errors Memory chips consist of a large array of tightly packed memory cells one for each bit of data The vast majority of the intermittent failures are a result of interaction between these memory cells Often writing a memory cell can cause one of the adjacent cells to be written with the same data An effective memory test attempts to test for this condition Therefore an ideal strategy for testing memory would be the following 1 write a cell with a zer
44. ing a problem Copyright 2015 Passmark Software Page 25 3 Troubleshooting Memory Errors Please be aware that not all errors reported by MemTest86 are due to bad memory The test implicitly tests the CPU L1 and L2 caches as well as the motherboard It is impossible for the test to determine what causes the failure to occur However most failures will be due to a problem with memory module When it is not the only option is to replace parts until the failure is corrected Sometimes memory errors show up due to component incompatibility A memory module may work fine in one system and not in another This is not uncommon and is a source of confusion In these situations the components are not necessarily bad but have marginal conditions that when combined with other components will cause errors Often the memory works in a different system or the vendor insists that it is good In these cases the memory is not necessarily bad but is not able to operate reliably at full speed Sometimes more conservative memory timings on the motherboard will correct these errors In other cases the only option is to replace the memory with better quality higher speed memory Don t buy cheap memory and expect it to work reliably On occasion block move test errors will occur even with name brand memory and a quality motherboard These errors are legitimate and should be corrected All valid memory errors should be corrected It is possible that a particular er
45. le parameters Improved and extended CPU identification routines Newer CPUID based method is now used to determine cache sizes for Intel CPUs for better accuracy and supportability Routines for calculating cache and memory speeds have been reworked for better accuracy An overflow problem has been fixed that resulted in no memory speed being reported for CPUs with large L3 caches Fixed some errors in the crash reporting routines Misc minor fixes and code cleanup Copyright 2015 Passmark Software Page 49 Enhancements in v4 0 28 Mar 2011 Full support for testing with multiple CPUs All tests except for 11 Bit Fade have been multithreaded A maximum of 16 CPUs will be used for testing CPU detection has been completely re written to use the brand ID string rather than the cumbersome difficult to maintain and often out of date CPUID family information All new processors will now be correctly identified without requiring code support All code related to controller identification PCI and DMI has been removed This may be a controversial decision and was not made lightly The following are justifications for the decision 1 Controller identification has nothing to do with actual testing ofmemory the core purpose of Memtest86 2 This code needed to be updated with every new chipset With the ever growing number of chipsets it is not possible to keep up with the changes The result is that new chipsets were more often tha
46. m data pattern and caching The execution order for these tests were arranged so that errors will be detected as rapidly as possible A description of each of the test sections follows Address test walking ones Tests all address bits in all memory banks by using a walking ones address pattern Errors from this test are not used to calculate BadRAM patterns Address test own address Each address is written with its own address and then is checked for consistency In theory previous tests should have caught any memory addressing problems This test should catch any addressing errors that somehow were not previously detected Moving inversions ones amp zeros This test uses the moving inversions algorithm with patterns of all ones and zeros Cache is enabled even though it interferes to some degree with the test algorithm With cache enabled this test does not take long and should quickly find all hard errors and some more subtle errors Copyright 2015 Passmark Software Page 34 Moving inversions 8 bit pattern This is the same as test 3 but uses a 8 bit wide pattern of walking ones and zeros This test will better detect subtle errors in wide memory chips A total of 20 data patterns are used Moving inversions random pattern This test uses the same algorithm as test 3 but the data pattern is a random number and it s complement This test is particularly effective in finding difficult to detect data sensitive errors A t
47. n addition the CPU must work properly to run MemTest86 so successful execution implicitly assures that your CPU is also functioning properly MemTest86 can not diagnose many types of PC failures For example a faulty CPU that causes Windows to crash will most likely just cause MemTest86 to crash in the same way See Troubleshooting Memory Errors for details on how to interpret memory errors detected by MemTest86 Copyright 2015 Passmark Software Page 20 2 4 3 1 Runtime Configuration Options MemTest86 may be configured during operation via runtime configuration commands Pressing the C key at anytime will display the runtime command menu When running the UEFI version the following options are available in the Configuration command menu Settings 1 Skip Current Test 2 End Test 3 Go Back to Main Menu 0 Continue The runtime configuration commands allow the user to adjust the following settings 1 Skip Current Test Aborts the current test and starts the next test in the sequence 2 End Test Stops the test and displays a summary of the results 3 Go Back to Main Menu Stops the test and returns to the main menu 4 Continue Resume the test 2 4 4 Test Report Pro version only At the end of the test a summary of the test results is displayed as well as the option to save an HTML test report to a file The test report is fully customizable by modifying the following files mt86head htm The HTML co
48. n not reported in correctly In the authors opinion incorrect information is worse than no information 3 Probing for chipset information carries the risk of making the program crash 4 The amount of code involved with controller identification was quite large making support more difficult Removing this code also had the unfortunate effect of removing reporting of correctable ECC errors The code to support ECC was hopelessly intertwined the controller identification code A fresh streamlined implementation of ECC reporting is planned for a future release A surprising number of conditions existed that potentially cause problems when testing more than 4 GB of memory Most if not all of these conditions have been identified and corrected A number of cases were corrected where not all of memory was being tested For most tests the last word of each test block was not tested In addition an error in the paging code was fixed that omitted from testing the last 256 bytes of each block above 2 GB The information display has been simplified and a number of details that were not relevant to testing were removed Memory speed reporting has been parallelized for more accurate reporting for multi channel memory controllers This is a major re write of the Memtest86 with a large number of minor bugfixes and substantial cleanup and re organization of the code Copyright 2015 Passmark Software Page 50 Enhancements in v3 5 3 Jan 2008
49. nd data that is valuable in diagnosing problems with test execution A large number of trace points have been added in key portions of the code in particular SMP startup routines to provide visibility of obscure failures This feature will allow non technical users to provide troubleshooting data for better test stability Added a new One Pass feature This feature runs the complete test once and then exits but only if there were no errors This provides a convenient method for unattended testing One Pass may be enabled via a boot option or via an on line command Images for CD USB key and Floppy disks now use Syslinux for booting and include a variety of standard options and two previous versions of Memtest86 The new boot time options may be specified at the boot prompt A feature has been added to allow customization of the list of tests to be run The test list may be specified via a boot option or via an on line command A feature has been added to restrict specific CPUs that are to be used for testing The maximum number of CPUs may be specified or a 32 bit CPU mask may be specified These are enabled with boot options A number of problem with use of on line commands when testing with more than one CPU have been fixed A selection of boot time parameters are were added These options enable boot tracing the One Pass feature limit the maximum number of CPUs to use specify a CPU mask to select CPUs to be used and setup serial conso
50. o 2 write all of the adjacent cells with a one one or more times 3 check that the first cell still has a zero It should be obvious that this strategy requires an exact knowledge of how the memory cells are laid out on the chip In addition there is a never ending number of possible chip layouts for different chip types and manufacturers making this strategy impractical However there are testing algorithms that can approximate this ideal strategy A 2 MemTest86 Test Algorithms MemTest86 uses two algorithms that provide a reasonable approximation of the ideal test strategy above The first of these strategies is called moving inversions The moving inversion test works as follows 1 Fill memory with a pattern 2 Starting at the lowest address 2a Check that the pattern has not changed 2b Write the patterns complement 2c Increment the address Repeat 2a 2c 3 Starting at the highest address 3a Check that the pattern has not changed 3b Write the patterns complement 3c Decrement the address Repeat 3a 3c Copyright 2015 Passmark Software Page 33 This algorithm is a good approximation of an ideal memory test but there are some limitations Most high density chips today store data 4 to 16 bits wide With chips that are more than one bit wide it is impossible to selectively read or write just one bit This means that we cannot guarantee that all adjacent cells have been tested for interaction In this case the best we can do is
51. o concurrently use multiple CPUs to test memory It may however be limited by the implementation in the underlying firmware Copyright 2015 Passmark Software Page 3 1 3 1 UEFI v5 For UEFI systems multiprocessor support is dependent on the implemention of the multiprocessor services provided by the UEFI firmware On older UEFI systems the multiprocessor support can be fairly limited causing issues such as a reduced number of CPUs available for testing or even program freeze when attempting to run on any CPU other than the first It is recommended that MemTest86 is run on only one CPU first before attempting to run on multiple CPUs 1 3 2 BIOS v4 For older systems that use the traditional BIOS MemTest86 will function properly with any number of CPUs but is currently configured to use a maximum of 32 CPUs for testing When runnining on 64 bit CPUs MemTest86 executes in 32 bit mode using PAE For 32 bit CPUs testing is limited to 64 GB 64 bit CPUs running MemTest86 executes in long mode which allows for testing of up to 8 TB of memory CPUs executing in 32 bit mode can test a maximum of 2 GB of memory at a time This 2 GB window is then advanced allowing for all of memory to be tested Copyright 2015 Passmark Software Page 4 2 Setup and Use MemTest86 supports booting from both the newer UEFI platform and the traditional BIOS When booting from UEFI MemTest86 has access to additional services not available in BIOS in
52. of detailed RAM SPD information Timings clock speeds vendor names and much more Support to writing to the USB drive that Memtest is running from for logging and report generation In all prior MemTest releases there was no disk support Use of GPT GUID Partition Table ECC RAM support limited hardware support ongoing development o Detection of ECC support in both the RAM and memory controller o Polling for ECC errors o Injection of ECC errors for test purposes limited hardware only Option to disable CPU caching for all tests Support for reading parameters from a configuration file to allow settings to be pre defined without the need for keyboard input This can help with automation Support for Secure Boot Speed improvements of between 10 and 30 Especially for tests 5 8 amp 9 This is the result more moving to native 64bit code removing the PAE paging hack switching compilers and using faster random number generation algorithms Addition of 2 new memory tests to take advantage of 64bit data and SIMD instructions Copyright 2015 Passmark Software Page 46 Enhancements in v4 3 7 May 2014 Fixed freeze particularly for older machines caused by incorrect handling of RSDP revision 0 in the multiprocessor detection code Added menu option for enabling serial console mode Enhancements in v4 3 6 Nov 2013 Fixed crash particularly for AMD machines that is seemingly resolved by adding CPU synchronication
53. otal of 60 patterns are used The random number sequence is different with each pass so multiple passes increase effectiveness Block move This test stresses memory by using block move movsl instructions and is based on Robert Redelmeier s burnBX test Memory is initialized with shifting patterns that are inverted every 8 bytes Then 4MB blocks of memory are moved around using the movsl instruction After the moves are completed the data patterns are checked Because the data is checked only after the memory moves are completed it is not possible to know where the error occurred The addresses reported are only for where the bad pattern was found Since the moves are constrained to an 8MB segment of memory the failing address will always be lest than 8MB away from the reported address Errors from this test are not used to calculate BadRAM patterns Moving inversions 32 bit pattern This is a variation of the moving inversions algorithm that shifts the data pattern left one bit for each successive address The starting bit position is shifted left for each pass To use all possible data patterns 32 passes are required This test is quite effective at detecting data sensitive errors but the execution time is long Random number sequence This test writes a series of random numbers into memory The initial pattern is checked and then complemented and checked again on the next iteration However unlike the moving inversions test writing and che
54. parallel mode When allocating memory for Bit Fade Test leave 1MB of free memory available to prevent firmware drivers from running out of memory Fixed potential crash or other unexpected behaviour due to memory issues with random functions Reports are now saved using UTF16 encoding to support Unicode characters Changed memory allocation behaviour by only pre allocating memory segments gt 16MB to prevent memory starvation When mapping memory layout removed several limits reducing the memory space tested Fixed memory being allocated after memory layout has been mapped thus changing the memory layout Fixed memory leak when cleaning up after test completion Fixed memory leak when decoding PNG files Fixed progress bar not displaying 0 on completion of a pass Updated to new UEFI SDK libraries UDK2014 Fixed memtest86v4 incorrectly booting to serial mode by default Various system info related updates fixes SPD CPU Enhancements in v5 1 0 May 2014 Fixed ECC error detection for Ivy Bridge E chipsets Fixed rounding of memory SPD timings On 32 bit systems systems with upper address limit gt 32 bits freezing during testing is now fixed Locking memory for testing is now more robust Added SPD support for VI8237S Intel X79 Intel NM10 Chipsets Fixed incorrect decoding of of banks in DDR2 FB SPD causing the memory stick size to be reported incorrectly Increased the number of supported memory modules from 16 to 32 In
55. ror will never show up in normal operation However operating with marginal memory is risky and can result in data loss and even disk corruption Even if there is no overt indication of problems you cannot assume that your system is unaffected Sometimes intermittent errors can cause problems that do not show up for a long time You can be sure that Murphy will get you if you know about a memory error and ignore it We are often asked about the reliability of errors reported by MemTest86 In the vast majority of cases errors reported by the test are valid There are some systems that cause MemTest86 to be confused about the size of memory and it will try to test non existent memory This will cause a large number of consecutive addresses to be reported as bad and generally there will be many bits in error If you have a relatively small number of failing addresses and only one or two bits in error you can be certain that the errors are valid Also intermittent errors are without exception valid Frequently memory vendors question if MemTest86 supports their particular memory type or a chipset MemTest86 is designed to work with all memory types and all chipsets MemTest86 cannot diagnose many types of PC failures For example a faulty CPU that causes Windows to crash will most likely just cause MemTest86 to crash in the same way 3 1 Hammer Test Test 13 Errors The Hammer Test is designed to detect RAM modules that are susceptible to disturbance e
56. rrors caused by charge leakage This phenomenon is characterized in the research paper Flipping Bits in Memory Without Accessing Them An Experimental Study of DRAM Disturbance Errors by Yoongu Kim et al According to the research a significant number of RAM modules manufactured 2010 or newer are affected by this defect In simple terms susceptible RAM modules can be subjected to disturbance errors when repeatedly accessing addresses in the same memory bank but different rows in a short period of time Errors occur when the repeated access causes charge loss in a memory cell before the cell contents can Copyright 2015 Passmark Software Page 26 be refreshed at the next DRAM refresh interval Starting from MemTest86 v6 2 the user may see a warning indicating that the RAM may be vulnerable to high frequency row hammer bit flips This warning appears when errors are detected during the first pass maximum hammer rate but no errors are detected during the second pass lower hammer rate See A 2 MemTest86 Test Algorithms for a description of the two passes that are performed during the Hammer Test Test 13 When performing the second pass address pairs are hammered only at the rate deemed as the maximum allowable by memory vendors 200K accesses per 64ms Once this rate is exceeded the integrity of memory contents may no longer be guaranteed If errors are detected in both passes errors are reported as normal The errors detected during Test 1
57. s such as the specific tests to execute address range to test and which CPU s are used in testing Copyright 2015 Passmark Software Page 9 2 4 1 Test Configuration 2 4 1 1 System Info The System Info screen displays the hardware information of the system it is running on including whether ECC detection is enabled MOI Ri AA n LAO Susten nfo BY CPU CPU Type Intel Core i7 2600 e 3 406Hz CPU Clock 3257 MHz wy Dest Selection CPU Temperature N A Logical Processors 1 L1 Cache 32K 94912 MB s Seem A ddress Range LZ Cache 6144K 31833 MB s L3 Cache N A PF Memory PU Selection Total Physical Memory 511M 14425 MB s Memory latency 26 416 ns 9 tart Test ECC Enabled View memory usage MAM Benchmark Memory caching Enabled fe Ste ttings save system information summary to file d E x it Use the Up Down arrow Keys to select an option MemTest8b W6 0 0 Pro View detailed RAM SPD info displays the SPD information stored in the individual RAM modules The RAM SPD information can be saved to a file on disk Pro version only Note Retrieving RAM SPD information is highly dependent on the system s motherboard CPU chipset Most common hardware should be supported but for some systems accessing the SPD modules using the mechanism required by the chipset may not be supported If this is the case please send the MemTest86 log file to helo passmark com along with details of your system
58. should not be of concern to most users Therefore a warning rather than an outright failure would ensure the user is aware of the issue and be able to take the necessary measures to mitigate the issue The details of the errors that were detected in the first pass of the Hammer Test but not during the second pass can be displayed in the report by specifying the configuration file parameter REPORTNUMWARN This parameter represents the maximum number of warnings to display in the report file New configuration file parameter AUTOMODE for full automation Enabling this parameter shall result in the following 1 Splash screen is skipped and the tests are started immediately 2 When the tests are completed the report is saved to disk automatically 3 System is rebooted Fixes Enhancements Shortened the test time for Hammer Test Test 13 by reducing the total number of hammers per row address pair Fixed issue with the main menu not displaying for certain EFI firmware Due to the fact that many EFI firmwares require the use of the obsolete ConsoleControl protocol to switch between graphics console mode the graphics console mode workaround is now enabled by default Fixed bug with details of empty RAM slots being displayed when using SMBIOS memory device information Copyright 2015 Passmark Software Page 40 Added Intel Skylake ECC support Updated Jedec manufacturer ID list for displaying the vendor name from RAM SPD Fixed Jedec m
59. sting O Memory caching disabled 1 Memory caching enabled PASS1FULL Specifies whether the first pass shall run the full or reduced test By default the first pass shall run a reduced test ie fewer iterations in order to detect the most obvious errors as soon as possible 0 Reduced test 1 Full test ADDR2CHBITS List of bit positions of a memory address to exclusive or XOR to determine which memory channel 0 or 1 is used This is useful if you know that the memory controller maps a particular address to a channel using this decoding scheme If this parameter is specified and MemTest86 detects a memory error the channel number will be calculated and displayed along with the faulting address Each bit position specified is separated by a comma For example ADDR2CHBITS 1 8 9 will XOR bits 1 8 9 of the address to determine the channel LANG Specifies one of the following languages to use en US English fr FR French es AR Spanish Latin American de DE German ja JP Japanese Zh CN Chinese Simplified zh HK Chinese Traditional REPORTNUMERRS Number of the most recent errors to display in the report file This number must be no more than 5000 REPORTNUMWARN Number of the most recent warnings to display in the report file This number must be no more than 5000 Currently this parameter is used only for the Hammer Test Test 13 AUTOMODE Spec
60. sure to record the reported memory bandwidth for each parameter change 6 If in step 4 the CPU hangs before memory errors appear then the CPU has less margin for error than the memory If available you may want to try reducing the CPU multiplier and then continue to increase the system clock until either the CPU stops or memory Copyright 2015 Passmark Software Page 31 errors occur This is helpful if memory timings are not adjustable or are ineffective 7 CPU and memory voltages are adjustable on some motherboards and increasing them may allow you to run at higher speeds In particular higher CPU voltages tend to be quite effective for over clocking However higher voltages also mean higher temperatures so be sure that you have plenty of case cooling and an effective CPU cooler Use carefully 8 Some motherboards allow you to use a system clock multiplier for memory The default is usually 1 1 or in other words the system and memory clock rates will be the same This setting is only useful when the memory and CPU operational limits are significantly different and can not be brought into balance using the techniques listed above Once you have established the operational limits of your system then you need to select settings that allow for a reasonable margin for error Do not be tempted to use the maximum settings A margin of 3 to 6 is recommended for reliable operation The easiest way to add operational margin is to simply reduce the
61. the CD burning software available on your system to create a CD ROM using the extracted ISO image Be sure that you create a CD image from the ISO file rather than placing a copy of the ISO file onto a data CD Look for Burn Image from File or similar option under the File menu of your CD burning software Create a boot able USB Flash drive 1 Download the Windows MemTest86 USB image zip file 2 Extract the contents of the zip file to a directory 3 4 Launch the ImageUSB application that was included in the zip file Plug in the USB drive Copyright 2015 Passmark Software Page 5 SS 2 ey Ol Select the USB drive from the list If it is not already selected select the image file included in the zip file Click Write to UFD After accepting a few more prompts this should give you a working bootable USB drive 2 2 Boot disk Creation using Linux Mac Create a boot able CD ROM 1 2 Download the Linux Mac MemTest86 ISO image Untar the package tar xvzf memtest86 iso tar gz An ISO image file and a README file will be created in the current directory Use the CD burning software available on your system to create a CD ROM using the ISO image Be sure that you create a CD image from the ISO file rather than placing a copy of the ISO file onto a data CD Look for Burn Image for File or similar option under the File menu of your CD burning software On a Mac you can use Disk Utility Create a boot able
62. then run at least a partial pass of MemTest86 Running at least 15 of tests 1 5 should be the minimum amount of testing for each iteration Continue increasing the clock rate until you get a failure Take your time and take good notes For each step be sure that you record the memory bandwidth reported by MemTest86 Some BIOS s automatically adjust memory timings according to clock rate You may find that by increasing the clock rate memory performance will decrease Once you find a failure back off on the clock rate until you find the point at which you get errors Once you find the point at which memory errors occur back the clock off one step and run a full pass of MemTest86 to confirm the operational limit In some cases the CPU will hang stop responding before memory errors are detected 5 Some motherboards will allow you to adjust memory timing Memory timings are typically listed as 4 values separated by hyphens However some motherboards only offer choices like fast faster and fastest There are two strategies for adjusting memory timing If memory errors are detected before the CPU exhibits problems then slower memory timing may be used to allow a higher system clock rate to be used The second strategy is to use faster memory timings to get more memory bandwidth without increasing the system clock rate It is impossible to know which values will effect errors Some of the memory timings will affect memory bandwidth and others will not Be
63. to be replaced When more than one module is installed you will need to determine which module is failing First consult your motherboard documentation to determine if memory modules may be used individually or if they must be used in pairs If your motherboard is configured with the minimum number of memory modules you will need to purchase a replacement memory Copyright 2015 Passmark Software Page 28 module in order to locate the failing one Using the guidelines in your motherboard manual to maintain a working configuration selectively remove modules from the system and rerun Mem Test86 to find the bad module s Be sure to carefully note which modules are in the system when the test passes and fails In most cases memory failures will be due to a faulty memory module and replacing the module will resolve the problem However the motherboard affects memory operation and may also cause memory errors There are instances where only a particular combination of memory and motherboard will cause errors This is typically due to the memory not being of sufficient quality and speed to keep up with the motherboard This memory may function properly when installed in a less demanding motherboard When errors are only detected by test 5 it is usually due to memory not being fast enough to work properly with the motherboard More conservative memory timing BIOS settings if supported may resolve these problems Otherwise higher quality memory may be req
64. trace point A total of 26 trace points are displayed and older trace points are lost as execution progresses Pressing any key will advance to the next trace point As initialization proceeds other portions of the screen will be filled in with information and the header lines will be erased Gathering Trace Information There are two types of failures where trace information is needed to help diagnose the problem hangs and crashes Gathering traces for a situation when the test hangs is very simple Just continue to press return until the test stops At minimum email the last 5 trace points Much more useful would be a digital picture of the screen For cases where the test crashes we need to see the trace points just before the crash This requires slowly stepping through the traces and identifying the point where the test reboots Then run the test again and stop at the trace point just before the crash and report the information Again email at least the last 5 trace points or better a picture of the screen Please email failure information to help Wdpassmark com Using Trace Information Technical users will be able to diagnose problems using trace data that previously would have required detailed understanding of Memtest86 s internal workings By following the trace points in the source you will be able to follow the path of execution and identify problems As needed new trace points may be added to the code to provide more detail when diagnos
65. tracted ISO image Boot from a disk partition using Grub 1 Copy memtest bin to a permanent location for example boot memtest bin 2 Add an entry in the Grub config file boot grub menu I Ist to boot memtest86 Only the title and kernel fields need to be specified The following is a sample Grub entry for booting memtest86 title MemTest86 linux16 boot memtest bin Boot from a disk partition using Lilo 1 Copy memtest bin to a permanent location ie boot memtest bin 2 Add an entry in the lilo config file usually etc lilo conf to boot memtest86 Only the image and label fields need to be specified The following is a sample Lilo entry for booting memtest86 image boot memtest bin label memtest86 3 As root type lilo Copyright 2015 Passmark Software Page 7 2 4 Using MemTest86 UEFI To start MemTest86 insert the CD ROM or USB flash drive into the appropriate drive and restart your computer Note The UEFI BIOS must be configured to boot from the device that MemTestg86 is installed on Most systems have an optional boot menu that is enabled by pressing a key at startup often ESC F9 F11 or F12 If available use the boot menu to select the correct drive You may see both the UEFI and BIOS as separate options Please consult your motherboard documentation for details On a Mac you need to hold down the c key while the computer is booting to boot from CD To boot from USB you need to hold down the A
66. u Settings 1 Test Selection 2 Address Range 3 Error Report Mode 4 CPU Selection Mode 5 Refresh Screen 6 Restart Test 7 Miscellaneous Options 0 Continue The runtime configuration commands allow the user to adjust the following settings 1 Test Selection Individual tests may be selected for execution or the current test may be skipped When a failure has been identified it is much faster to troubleshoot by running only the failing test 2 Address Range With this option memory testing may be restricted to a selected portion of memory This is also useful to speed up the troubleshooting process by testing only the failing portion of memory 3 Error Report Mode The error report mode may be changed to provide different types of error information The default Error Summary mode provides all of the details required for diagnosing memory problems However other error reporting modes are available BadRAM patterns may be used with Linux systems to map out bad memory blocks 4 CPU Selection Mode This option controls how CPUs are selected for testing The options are All Round Robin and Sequential With Round Robin CPUs are selected in round robin fashion for each test With the Sequential option all available CPUs are Copyright 2015 Passmark Software Page 23 used for each test 5 Refresh Screen Redraws the screen when the display becomes garbled 6 Restart Test Restarts test wit
67. uired or possibly the motherboard may need to be replaced 4 4 Error Validity There are many users who question if errors detected by MemTest86 are valid In at least 99 9 of cases reported errors will be legitimate and must be corrected MemTest86 simply exercises all available RAM looking for errors If any error is detected in RAM regardless of how or where it is a legitimate failure that needs to be corrected There are no known issues regarding compatibility It is possible but unlikely that a particular error will never show up in normal operation However operating with marginal memory is risky and can result in data loss and disk corruption Even if there is no overt indication of problems you cannot assume that your system will be unaffected There are some rare cases where motherboard manufacturers will map hardware registers into space normally occupied by memory When these locations are tested errors are reported In this case the reported errors will be invalid To better identify these rare cases where invalid errors are reported an error confidence value is created by MemTest86 The program tracks failures and then does some simple analysis to determine the probability of invalid errors When the confidence value exceeds 100 it is nearly impossible that the reported errors will be invalid Copyright 2015 Passmark Software Page 29 5 Over Clocking 5 1 Background MemTest86 is an invaluable tool for over clocking and ma
68. y and CPU Using the guidelines below experiment with the settings available for your motherboard to find settings that result in the highest clock rate combined with the highest reported memory bandwidth Copyright 2015 Passmark Software Page 30 1 Before attempting to over clock you need to know what system parameters your motherboard will allow you to adjust and how they are adjusted Some motherboards will allow all useful parameters to be adjusted while others do not allow for any adjustment Consult your motherboard manual for details Some of the parameters that may be available are e System clock rate e CPU clock multiplier e Memory timing e Memory clock multiplier e CPU voltage e Memory voltage 2 You need to know how to reset the CMOS to factory settings It is common when over clocking to end up with settings that will not run the BIOS When the parameters are set via CMOS the only way to recover is to reset the CMOS to the factory configuration For some motherboards this is accomplished by removing a jumper Other will require removal of the CMOS backup battery Make sure that you Know how to recover before starting 3 Before adjusting any parameters run MemTest86 for a full pass to establish a baseline Record the memory bandwidth and CPU clock frequency for the default configuration 4 Typically the best place to start is with the system clock rate Increase the clock rate in fairly small increments 2 4 MHz and
69. y be used to increase your systems performance and reliability Many shy away from over clocking fearing that it will make their PC less reliable With a proper procedure fine tuning your system timings is safe and in some cases will even improve reliability Before embarking on over clocking one must understand the concept of margins or margin for error 5 2 Operating Margins To achieve high reliability computer systems are designed and tested under conditions that are more strenuous than those expected in normal use Take for example a computer that is designed to operate with a bus frequency of 133 MHz A quality manufacturer will design and test for operation at perhaps 140 MHz or higher This margin for error provides confidence that even with inevitable manufacturing variances and changing conditions operation will be reliable When components from different manufacturers are combined an even greater margin for error is required since the exact characteristics of the associated components are not known The result is the majority of computers will end up having a much larger margin for error than is needed This means that a lot of performance is being wasted On the other hand there will be a small number of systems that do not have enough margin of error and will have poor reliability even without over clocking Many think of over clocking as simply making a computer operate as fast as possible This approach is unwise and will often result
70. zing the screen for some firmwares due to an unsupported driver protocol Added workaround for incorrect text length height returned by the UEFI firmware Added checks for the number of cores exceeding the maximum supported in MemTest86 Copyright 2015 Passmark Software Page 41 Synchronized cache enabling disabling across all CPUs Migrated CPU cache info code from Performance Test Burnin Test The displayed cache info should now be more consistent with what is displayed in Burnin Test Performance Test Fixed Enabling Disabling of features in the Sys Info screen to be less confusing Fixed CPU Selection screen to truncate the list of available processors when more than 16 are available Various system info related updates fixes CPU Changes in v6 0 0 Feb 2015 New Features Support for DDR4 RAM and associated hardware including retrieval and reporting of DDR4 specific SPD details This includes DDR4 RAM that support Intel XMP 2 0 DDR4 RAM timings New RAM benchmarking feature allowing results to be graphed and saved to disk Previous results can be graphed on the same chart for comparison New Hammer Test for detecting disturbance errors caused by charge leakage when repeatedly accessing addresses in the same memory bank but different rows in a short period of time Language support for French German Japanese Chinese All text are displayed in the selected language including generated reports Fixes Enhancements Added Hasw

MemTest-86 User Manual Version 6.2

Contents

Download Pdf Manuals

Related Search

Related Contents