Memtest86 discussion

[eva2000]

note: new improved version of memtest86+ v1.20 available at www.memtest.org with a 12th test added



*** Enhancement in v1.20 : ***

New Features
Added advanced support for Intel Grantsdale (i915)
Added advanced support for Intel Alderwood (i925)
Added advanced support for VIA K8T800 Pro
Added support for Prescott core E0
Added support for Cyrix III CPU
Added support for VIA C3 Samuel2/Ezra/Nehemiah CPU
Added support for .5 FID for AMD K8
Added detection for VIA CLE266 and PLE133 Chipset


Bug Fixes
Complete rewrite of the Cool 'n Quiet core
Fixed a bug with K8 > 2.2 GHz
Fixed a bug with test #12 (no kb input check)
Removed the KTxxx (K7) advanced detection completely
Removed Bit Fade Test from the test batch
(Can still be run standalone)
Fixed a bug with VIA VT82C693A/694X detection
New final versioning scheme used for *nix users




...........................
Well since it has been brought up so many times i thought i'd post a reply I did to a post at asusboards.com about my thoughts on memtest86 and how i test :)

General

memtest ain't 100% but you can use memtest to guage the max possible FSB/MEM which is the top limit of what you can expect... since i don't think i've ever experienced windows 100% stability and error free at a speed higher than the highest memtest passable speed

therefore

max FSB/MEM speed (100% windows error free/stability) <= max FSB/MEM speed (memtest error free)

My experiences and testing methods

test #1 - 4
are cpu fsb speed and or vcore related (meaning lowering fsb or increasing vcore saw errors in these tests disappear)

test #5
prior to 865/875 boards - memory speed, timings and vdimm related (meaning altering mem speed, timings and/or vdimm saw errors in this test disappear)

test#6
with 865/875 boards bigtoe has said related to cycle time (tras) in cpuz which i sort of confirmed with my current testing

note: prior to 865/875 boards, i've never had memory errors in test #6 only since these new boards have i experienced test #6 errors

test #7
not sure very rarely have i experienced errors

extended test
test #8 is a more intensive version of test #5

most memory related errors pop up at test #5 hence i like looping test #5 for memory testing for 12-24hrs after 1-4hr general standard loop of test #1-7

i like to loop test #3 and/or #4 for cpu related issues

I follow it up with at least goldmem 5.07:

2 quick test loops
+
2-4 full standard loops <-- can take forever trying doing it with 4 x 512mb xms3200c2

to loop a particular test


in memtest hit keyboard key

C -> 2 -> 5 and then select a test number to loop

spacebar = locks screen so it shows the very first error so where ram first errors out

enter key = unlocks screen


this is done, before i even load windows... if theres errors in above stage memtest or goldmem stages, i don't even bother loading windows or benching

i realise not everyone has time to do this but i like to know exactly what i get (ram) and how it performs

some people just load windows and see if it crashes or is stable... but you can't see why it crashes or isn't stable - is it the memory or cpu related etc

it makes it easier to test when you have more than 1 pc and a kvm switch to control 2-4 pcs from the same console

update: Aug 10, 2003

Interesting while reading up on memtest86 I found something i didn't catch before, the description of test #4 http://memtest86.com/#details



Test 4 [Moving inv, 32 bit pat, cached]

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 effective in detecting data sensitive errors in "wide" memory chips.


so if you're running standard test #1-7 the bare minimum passes you're want to run would be = 32 passes to test all data patterns

For 256MB stick to do 32 standard test passess of test #1-7 would take between 75-160 minutes - longer for more memory

update: June 1, 2004: - added some snap shots for memtest86+ v1.11

here's some snapshots of memtest86+ v1.1 in action on various boards i tested

http://www.fileshosts.com/memory/memtest86_plus/examples1/web/memtest_000_thumb.jpg (http://www.fileshosts.com/memory/memtest86_plus/examples1/web/memtest_000.html) http://www.fileshosts.com/memory/memtest86_plus/examples1/web/memtest_001_thumb.jpg (http://www.fileshosts.com/memory/memtest86_plus/examples1/web/memtest_001.html) http://www.fileshosts.com/memory/memtest86_plus/examples1/web/memtest_002_thumb.jpg (http://www.fileshosts.com/memory/memtest86_plus/examples1/web/memtest_002.html) http://www.fileshosts.com/memory/memtest86_plus/examples1/web/memtest_003_thumb.jpg (http://www.fileshosts.com/memory/memtest86_plus/examples1/web/memtest_003.html) http://www.fileshosts.com/memory/memtest86_plus/examples1/web/memtest_004_thumb.jpg (http://www.fileshosts.com/memory/memtest86_plus/examples1/web/memtest_004.html) http://www.fileshosts.com/memory/memtest86_plus/examples1/web/memtest_005_thumb.jpg (http://www.fileshosts.com/memory/memtest86_plus/examples1/web/memtest_005.html)

[Hawk]

Nice info Eva2000. . . That should help alot of people out. . . .

:wave:

[Mikki]

I can confirm a lot of eva2000's results...and I've also noticed from other's results that the difference when running newer chipsets is there. I'm still on GB.

Nice thread mate! :-) :beer:

[eva2000]

glad you guys like it :)

[monopoly]

Thanks for the test list info, eva2000. Your observations appear to be on mark with error/corrections - and which test relates to which setting area.

Much help to me.

[eva2000]

Originally posted by monopoly
Thanks for the test list info, eva2000. Your observations appear to be on mark with error/corrections - and which test relates to which setting area.

Much help to me. good to hear

for anyone interested ryan/ocguy at Asusboards maybe posting some more info re: memory testing and error finding http://asusboards.com/forums/showthread.php?s=&threadid=50588

[Chuck232]

Nice eva2000! Very useful info! :D

[eva2000]

anyone notice when oc'ing 1:1 that sometimes memtest86 L1 and L2 cache reported speeds match each other at certain speeds and timings and sometimes L1 is > L2 reported bandwidth ? could it be at certain oc's if not enough vcore, vdimm, L1/L2 and mem reported bandwidth is inaccurate/messed up ?

[eva2000]

Interesting while reading up on memtest86 I found something i didn't catch before, the description of test #4 http://memtest86.com/#details



Test 4 [Moving inv, 32 bit pat, cached]

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 effective in detecting data sensitive errors in "wide" memory chips.


so if you're running standard test #1-7 the bare minimum passes you're want to run would be = 32 passes to test all data patterns

For 256MB stick to do 32 standard test passess of test #1-7 would take between 75-160 minutes - longer for more memory

[ThugsRook]

For 256MB stick to do 32 standard test passess of test #1-7 would take between 75-160 minutes - longer for more memory
not on my system it dont, and ive got 1024mb :-r

hmm i wonder exactly how long it would take?
i guess ill go see ;) brb...

[eva2000]

Originally posted by ThugsRook
not on my system it dont, and ive got 1024mb :-r

hmm i wonder exactly how long it would take?
i guess ill go see ;) brb... for 1GB IIRC would take 30-50mins per standard loop, i remember for 2GB standard loops took well over 60mins each LOL

time taken increases expontentially with more memory :(

[ThugsRook]

1 pass (#1-7) took 7.5 mins with 1024mb
32 passes will take 60 mins with 256mb
32 passes will take 240 mins with 1024mb

not bad :)

[eva2000]

Originally posted by ThugsRook
1 pass (#1-7) took 7.5 mins with 1024mb
32 passes will take 60 mins with 256mb
32 passes will take 240 mins with 1024mb

not bad :) ah got my wires crossed, was think goldmem 5.07 dos test, 1 loop would take 50+ mins hehe

[eva2000]

interesting test #10 needs to be added to the list of tests to loop for errors..

right now it seems best to start with

1. two full loops of all tests #1-11
2. looping of tests number 3, 4, 5, 6, 8, 10 for min 32-64 loops but better to loop for 6-24hrs

[ThugsRook]

hehe~
just loop all 11 for 10 loops and be done with it ;)

:wave:

[quick]

hi,
i'm new to using memtest86. I have a cpule of quesitons.

1. What do you mean by looping? is that running the test for a second time?
2. how long do you think i should run memtest86 to get an extrememly accurate analysis of my ram??
3. Is the only memtest you can run, the one that u use a floppy and reboot with it and go into it? thus you can't do anything else on the computer while its running.


my system:
2.6C
Ic7-max3
ocz 4200 el (1 gig)
sp94 w/ tornado
2x wd raptors
9800pro

THANKS.

[Jim]

Welcome Quick . :wave:

Memtest will continue to loop (run all of it's tests over and over) until you hit escape, remove the floppy and boot into your OS.

How many times? Personal preference I guess. I'm happy with an hour or two error free. :)

[quick]

so after i boot up into memtest, i can take out the floppy and it will continue to do memtest?

[quick]

after i take out the floppy it will boot into widows and at the same time it will do memtest is what i meant to ask.

[Ralf Hutter]

Originally posted by quick
after i take out the floppy it will boot into widows and at the same time it will do memtest is what i meant to ask.

No it won't boot into Windows, it'll keep running Memtest86 until you hit the "escape" key, then it'll boot into Windows.

[Ardy]

Interesting,I put memtest on cd and it boots to it and runs 1-7 tests on a loop,but trying to exit or use config,nothing happens??

[Ice]

Great information, eva. Thanks!

:)

[eva2000]

to fix v1.10 bugs

http://www.memtest.org


Enhancement in v1.11 : *

New Features
Added detection for Pentium-M 0.13 & 0.09 µm (Banias & Dothan)
Added detection for Pentium 4-M 0.09µm (Prescott)
Added detection for Centrino platform
Added complete advanced settings detection for i852 and i855 series (Mobile)
Added experimental detection for ALi M1687, AMD8000 and Intel i7505
Added detection for VIA KM266/KN266/KM400 and SiS645/645DX/648/748
A newsletter for memtest86+ update is now available. Check www.memtest.org

Bug Fixes
Instant-Reboot and crash with Prescott CPUs solved
Channels detection code for nForce2 rewritten.
Thoroughbred B and Duron Applebred now detected as 0.13µm
Added i848P to springdale's string.
Pass Count now resetted when selecting individual test
Now compile again with gcc2
Lots of others Bug Fixes

[user1]

is that a sword or a whip you are playing with there...?
not that I am in to that sort of thing, just, ahh curious. :cool:

[eva2000]

a Japanese sword of course :D

[eva2000]

......

[eva2000]

added some snapshots of memtest results on various boards to first post in this thread http://www.bleedinedge.com/forum/showthread.php?p=5598#post5598 enjoy :)

[Subsider]

i can run memtest + looped for ever at 249 mhz (ddr 498) at 250 (ddr500) ill get 1-3 random errors in test 5 and/or test 8 after 8 - 24 hrs of testing . Tony from OCZ says that the memory controler would take care of these random errors and it shouldnt be anthing to worry about . what do you guys thing ???? also vdimm is set @ 2.65 if i raise the vdimm any the errors become more numerous and come much sooner it testing !

[eva2000]

i can run memtest + looped for ever at 249 mhz (ddr 498) at 250 (ddr500) ill get 1-3 random errors in test 5 and/or test 8 after 8 - 24 hrs of testing . Tony from OCZ says that the memory controler would take care of these random errors and it shouldnt be anthing to worry about . what do you guys thing ???? also vdimm is set @ 2.65 if i raise the vdimm any the errors become more numerous and come much sooner it testing ! sounds like NB being stressed there... but loop memtest86+ test #5 for a while and see.. i never run windows when memtest errors are present that's just me.

For everyday use not benching 245mhz stable would still be better than 250mhz semi-stable as you wouldn't really notice a difference in performance

[Subsider]

eva
got it now just need alittle tweakin (ram timmings mainly) and a better fan on the NB i should do somthing to cool my ram though . ive ran memtest 14-16 hrs with no errors test 5 , 8 and full range .been @ 250mhz for over a week total stable

[KaEL]

Is just me, or Athlon64 FX on test 7 (memtest+ 1.30) takes more time than P4Cs,

have you guys notice?

Is it normal? I wonder if is due to the A64s memory controller

[RubberDuck]

Thx eva2000 That Cleared It Up. That Is Usful Info
:bow: :bow: :bow: :bow: :bow: :bow: :bow:


AMD Is Like A Lamborghini & Intel Is Like A Ford Pinto. AMD Is Awesome & Intel Is Just Merely Chopped Liver.( RubberDuck )

[nitewave]

test#6
with 865/875 boards bigtoe has said related to cycle time (tras) in cpuz which i sort of confirmed with my current testing

note: prior to 865/875 boards, i've never had memory errors in test #6 only since these new boards have i experienced test #6 errors


I'm seeing errors in this test (Abit IS7, i865PE), but what in the heck am I supposed to do about it? Mess around with cycle time until I don't see the errors?


Paul Braman
nitewave<at>tampabay<dot>rr<dot>com

[timpanogos]

Memtest386 V1.51 (nF4 embedded)
L1 Cache 128k 21522 mb/s
L2 Cache 1024k 5347 mb/s
memory 1024m 2564 mb/s

250mhz (DDR500) 2-2-2-6

32 passes of test 1-8 --- WallTime: 09:23:54

[gijoe]

Idd nice post,
Bit late into this thread so i hope some peeps are still following?
The reason i am late is because i just really started to try to push my ram(see sig) and it doesnt really seem to want to go anywhere above the 270mark.started this thread today
http://www.bleedinedge.com/forum/showthread.php?t=9865
So my Q as i have the dfi i can run this direct from the bios which i have done in the past,now just to throw some eg in;
I am a neebie at this :( so here goes
Say just throw caution to the wind and set up at 10x280(knowing that anything under 2900 seems to be stable on this rig) say it pass after teasting ,then i change to10x285 but it fails,How do i know if its cpu or ram related?
I suppose to some these are typical neebie question but hey have to start somewhere.
If someone can help explain where to see the differences i think ill be able to find my max for both cpu and ram.

Its at the horrible stage of not getting anywhere as id like to try to get stable at 3gz which i thought possible with this rig but now i am really not so sure.
As well i see guys hitting over the 300 marks stable which is tbh damn anoying (lool)
any help is well appricated
best
joe

[BB_One]

Simple gijoe,

Set your cpu say at cpu multi of x7, now start increasing and testing your FSB. Increase fsb until ram errors out in mentest and tweak until you know you have reached the max. This will tell you until when the mem is stable and takes the cpu out of the equation.

Once done and you found your max FSB for the ram, start increasing your cpu multis. If you have reached for example 250 max memory and u thingk your cpu still has some room to move, set a divider for your ram say 9/10 (188) for example this will reduce the fsb of the ram but not the system hence you can keep increasing your fsb and cpu multi.

[gijoe]

lol always sounds simple m8 :)
first change ill see how it goings then ill be back to haunt you :)
thx
joe

[rok]

nice post eva........BTW, I thought test 7 was an uncached fine tune?

[bugmenot]

try the latest memtest86 released today at http://www.softwarecove.com/memtest/

[dcloud]

That's all Greek to me. I downloaded MemTest because it was recommended by someone on another forum, but I haven't a clue what to do with it. I extracted the zip file into a folder (which I named MemTest) and I read the ReadMe file:

To build Memtest86:
1) Review the Makefile and adjust options as needed.
2) Type "make"

This creates a file named "memtest.bin" which is a bootable image. This
image file may be copied to a floppy disk or lilo may be used to boot this
image from a hard disk partition.

To create a Memtest86 bootdisk
1) Insert a blank write enabled floppy disk.
2) As root, Type "make install"

So I ran the MemTest .exe file and a command window came up and I typed "make install" but nothing happened. I tried again typing "A:> make install" and nothing happened.

What would be great is if someone could include instructions for using this thing for uber n00bs like me :)

- dc

[ketamine]

That's all Greek to me. I downloaded MemTest because it was recommended by someone on another forum, but I haven't a clue what to do with it. I extracted the zip file into a folder (which I named MemTest) and I read the ReadMe file:

To build Memtest86:
1) Review the Makefile and adjust options as needed.
2) Type "make"

This creates a file named "memtest.bin" which is a bootable image. This
image file may be copied to a floppy disk or lilo may be used to boot this
image from a hard disk partition.

To create a Memtest86 bootdisk
1) Insert a blank write enabled floppy disk.
2) As root, Type "make install"

So I ran the MemTest .exe file and a command window came up and I typed "make install" but nothing happened. I tried again typing "A:> make install" and nothing happened.

What would be great is if someone could include instructions for using this thing for uber n00bs like me :)

- dc


I believe those instructions are for Linux. Try downloading the the bootable .iso file for windows. Burn-it, leave it in the drive, set your boot (in BIOS) to CD-ROM drive and you should be good.

[ryant]

does it really makes a beep sound when it gets t o errors ,on error mode?
iam talking about version 1.65 ....thanks.

[galaxyconvoy]

I just got new memory in today and decided to see just how much I can get out of this. So I'm hoping to get answers to these questions I have.

I've used the original Memtest86 first and so I have grown used to it. Looking back at the original post and the subsequent update(s), I found that the "Test 4" that was noted is now Test 6 in the original. Would using the original Memtest86 be thorough enough to detect any hint of errors if I were to say loop this particular test many times? I've 2x1GB and would like to know if this'll be enough to ensure the speeds I test are good enough for everyday use.

Also, what is/are the advantage(s) Memtest86+ has over the original that warrants the need to use it over the original in terms of testing for errors and so forth?

[BGP Spook]

Well to start with memtest86+ has some OCing options. (Not sure about the original.) You can use if you have one of the supported chipsets and motherboards.

The newest versions have support for more models of motherboards and CPUs.

Also the error detection is, I believe, slightly better over the original. But, this is a subjective argument.

[galaxyconvoy]

Much thanks for the reply. I'll give the + version a try after I'm done with a few other tasks. I read through some of the things implemented in the + version and it looks interesting. Never gave much thought into using the + over the original. Maybe after a few test runs I'll be convinced to convert.

[galaxyconvoy]

More questions! :) This one is probably a FAQ too. It was mentioned that Test #5 would test whether the new memory speed, timings, and vdimm voltage are stable. Is a single pass of this particular test enough to determine if the new settings is considered stable enough for everyday use? Or does one need to loop the test many times (or 32 times?) to determine that? (If the O.P. is outdated in which test is which, which one do I run to test for this?)

I just got done running test 1-8 for 16 loops straight. Feel free to use some of these documented info to better gauge how long the process takes.

Memtest86+ v1.65
Total time: 10h39m
Tests: 1-8
Estimated time per 1-8 run: 39 mins 56.4 secs
Estimated time for 32x run: 21 hours 18 mins
CPU: Athlon64 X2 3800+ @ 2.4GHz (240x10) 1.40v (+0.075v adjustment)
Memory: Gigaram PC3200 2GB kit (rated 2.5-4-4-7 @ DDR400), 3-4-4-7 @ DDR480

[niksoftware]

I can pass memtest for many loops fine at all sorts of overclocks

But I find it incredibly difficult to make any overclocks stable in prime95


is there particular settings which tend to affect windows+Prime95 that memtest can cope with?

eg volts more important coz of the genereal windows drain or a particular timing?


i can do 2.5-3-3-7 @ 270 with 2.5v(reads 2.55v) in memtest fine
and 2.5/4/3/7 @ 300 with 2.8 or so volts(prolly less now) fine in memtest too

It is hard enough twidllin timings and volts to get it not to bsod in windows nevermind prime95 stable.

It makes ocing so slow its infuriating help plz

[ol'norton]

I just run Memtest V1.65 for the first time, but I'm not sure exactly what was tested. I just put the disk in and let it go while I was at work. When I left it was testing L1 Cache. When I got home and watched it for a while It was Testing L1 Cache. Did it test the other memory while I was away? Or do you have to configer it? I was going to include the screen shot I took but I couldn't figure out how to.:coocoo:

[Timber Jockey]

Ya if you just let it rur it tested the memory you have. It runs a serise of tests tests 1-8. So if there were no errors you are golden. But even passing memtest dosen't mean you are stable in windows. Only Prime 95 or Stress prime will tell you if you are Windows Stable.

[jaz]

I'm not that new with memtest86 and i still can't get the dam thing to work. I have it on floppy and burned succesfully TO a dvd. Both boot up and start the tests automatically. but i have no control over the menu, besided making it appear and vanish.
Really frustrated with this because i want to get on with my 'ocing, but i need memtest to ensure stability.
Can anyone PLEASE help?

[jaz]

pretty please with sugar on top?? :smash:

[RyderOCZ]

Jaz,

What do you mean you don't have control?

I take that to mean when you press C that nothing happens? You don't get the menu?

What type of Keyboard do you have?

[jaz]

It's a ps2 keyboard. I disable usb, (on dfi nf3, usb MUST be disabled or it doesn't work right. My mouse is usb, and i have a usb digital camera attached, but like i said, i disable usb support when i run it. I have tried using it with my old asus k8v deluxe, but with same rusults. Ive tried having nothing usb plugged in and usb disabled on both boards. My dfi ut nf3 250Gb has a built in memtest86 (V1.55). And i have that disabled, or it will try to run both versions of memtest and I don't want to see what happens if i allow that to proceed.

But i have V1.65 (latest version) burnt TO a dvd (not ON it, successfully, and i have it on floppy also. I even re-downloaded it and saved it to a freshly formatted floppy.

When i go into bios and enable boot-up floppy, the pc reboots into memtest and memtest automatically starts running. It work fine, running through all the tests, no problem.
But when i try to use any of the commands, they don't respond. I hit "c" to summon the menu, i can see it (black with white text) and the options numbered, but i hit "c" 1-3-5-enter-0 like i see has been posted all over the place to choose a test to loop, but it just resumes after it's pause when i have the menu up. I tried holding down the "c" key and all kinds of numer combinations and it just refuses to respond.

I am using the memory brand new too, in my sig and it runs all the way through error free when i just set them to cpu-z's settings as i see them. Then i got a64info (latest version) wrote down the settings as i saw them, but used the bios to enter the settings (not a64 info) I don't use ANY 'oc third party software.

Now I'm getting like 60 errors near the end of test #5. I up the juice to 1.7V from 1.6V
and I'm down to two errors. but the fact remains, i have to wait forever until it gets to test #5 (the one i need the most for 'ocing and testing) and test 8.

Once i can pass these two individual tests, then i will run overnight for total stability test. That's what most people do as Ive read on many forums, including the DFI forums.

And I'm not even 'ocing yet.

[ol'norton]

I thought it was only Legacy USB that had to be disabled?

[jaz]

I'm not sure, but the dfi forums overclocking guide says to disable usb to use memtest or it won't run properly or something to that effect. I forgot once to turn off usb support completely when i ran memtest once and got about 600 errors. I turned it off and it was down to about 60 errors, like i said in a previous post.

[jaz]

anybody else having problems with memtest86?

[jaz]

I thought it was only Legacy USB that had to be disabled?

I don't have a "legacy" option. It's just usb supprt in general on the dfi's

[Just Learnin']

You posted in another thread that you were unsure how healthy your floppy drive was. Have you tried running it from a CD with the questionable floppy drive disconnected? :scratch:

[jaz]

I can pass memtest for many loops fine at all sorts of overclocks

But I find it incredibly difficult to make any overclocks stable in prime95


is there particular settings which tend to affect windows+Prime95 that memtest can cope with?

eg volts more important coz of the genereal windows drain or a particular timing?


i can do 2.5-3-3-7 @ 270 with 2.5v(reads 2.55v) in memtest fine
and 2.5/4/3/7 @ 300 with 2.8 or so volts(prolly less now) fine in memtest too

It is hard enough twidllin timings and volts to get it not to bsod in windows nevermind prime95 stable.

It makes ocing so slow its infuriating help plz

I had to replace my antec true430 because of power fluctuations my dfi board was picking up on. These dfi boards are real fussy when it comes to power supplies and memory, (of which yours is fine), but i was ****ed.

I paid $100 for that power supply. It served my dumba** asus board with it's mixed memory well, but when i upgraded to the dfi board, well you know the rest. You'll probably always get strange errors and never get rock solid stable with that psu. If i were you, i would pay a visit to your mobo forum and look at the recommended psu's for your board. See what it says about antec.

I'm not bashing antec, but it seems they are having some problems with psu's like yours (and my old one which will be in my other pc with my numb asus board that tolerates mixed memory)

[jaz]

You posted in another thread that you were unsure how healthy your floppy drive was. Have you tried running it from a CD with the questionable floppy drive disconnected? :scratch:

No...i haven't. But I thought a floppy drive needs to be attached to the pc or it won't boot or something?

[Just Learnin']

There is a downloadable ISO as well for the CD so....? Not sure if runs from the CD or ??? You should not need the floppy drive attached to the PC for the rig to boot. Just change your boot options in the BIOS.

[jaz]

I have memtest burnt correctly TO a dvd, not on it, like some people make the easy mistake of doing. I'm going to replace my floppy drive with my old dell drive (works fine) anyway, so i will try what you say while it's not connected first.
Thanks :thumb:

[Just Learnin']

I hope it works! You have been trying to get this one beat for a really long time. :yikes:

[jaz]

Please, don't remind me. :smash:

[J4bb3rw0ckY]

Finally a new version of memtest86+ arrived. Can be downloaded at www.memtest.org.
It's much better now and supports latest CPU's and chipsets.

But I can't still determine my prob :cry:
Mobo says I have 1023Mb useable RAM of 1024MB and when testing memsizing it displays about 60 errors after 1022.1 Mb.
I think I will need some other testing software too I guess :smash:

[liquidcooled]

ver 1.70 dont show my chipset info, but does show L2 cache, and cpu type.....but nothing else.

[wpshooter]

//

[cye]

Hi,

I am new to memtest86, i do some changes on the coding so that only run for one loop of testing cycle..but i don't know what is the suitable compiler to run it and how can execute it.

I need to compile memtest86 into .exe file.

I did try codeblock, but error shows that no sys/io.h file exists.

Please let me know how to use code block to compile, or there is any other compiler that i can use.

Thanks very much

Have a nice day

Best Regards,
CYE