Memtest86 V3.4

[MadScientist]

I downloaded the iso file and the file for a floppy boot disk of the latest version of Memtest86 V3.4. I burned the iso with Nero. I tried this on 2 different computers with the same results. Neither the CD or the floppy worked. It starts but almost immediately I get an "Unexpected Interrupt-Halting". I tried the same with V3.3 using the same CD-RW and floppy disk and it works fine. Anyone get V3.4 to run?

I'm currently using Memtest86+ V1.7 which works.

 

[BonzaiDuck]

I can't tell you how to solve your MEMTEST86 3.4 problem. Believe me, I have part of those troubles, too. I lean toward using the floppy over the CD-burn, but still pray like hell that making the floppy results in a useable copy. I've discovered also that when you make the floppy on one computer and confirm that it works, there's some chance that it won't work on another computer. It COULD be the case, though, that if you try burning the floppy on another computer, it will work with yours. Don't ask me why, although I think I recall such happening to me once.

If it were only possible to put the MEMTEST86 program on a USB flash drive, it would help enormously. But the program makes a floppy unreadable by DOS or windows, so it may either be impossible to do it with a flash/thumb drive, there are some particulars we don't know, or you risk damaging the flash-drive (somehow).

But here's my question -- which I append to yours, if only to use your thread parasitically to troll for answers.

Some people have remarked that MEMTEST86 (this or that version -- or even all) doesn't play "nice" with 680i chipsets. On my system it "works," but it doesn't post information about the chipset, and it reports erroneous bandwidth estimates for the RAM.

Does anyone have more detailed information about this, or whether there is a version that works with nVidia chipsets with the same accuracy and relability that it works with the Intel chipsets?

 

[MadScientist]

A lot of rhetoric Duck but a question is not an answer. The 2 motherboards I tried V3.4 on have an nvidia nforce4 chipset and an Intel P35 chipset. Burn an iso image of V3.4 to a cd and see if it works on your 680i. I'm really just curious if V3.4 works or not, since Memtest86+ V1.7, it's equivalent, works perfectly.

 

[BonzaiDuck]

Well, sorry I didn't solve your problem -- the thread was a good place for me to sound off, since this has just become an issue with me.

First of all, I notice that the purveyors' of MEMTEST86 v 3.4 are selling the disc for $10.

I will try and find time to burn the CD in the next few days.

But think about that. You can get a CD-R for pennies, and you can run a computer like a factory burning discs. How convenient that the ISO download doesn't make "bootable" when you download it for free!! You may be better off with the floppy. Or -- give them the $10 (and if it doesn't work then, you REALLY have a gripe to make with them.)

I've seen some very established, reputable software firms with products that seem to exhibit "failures" just before _ say -- an AV "virus profile subscription" is about to expire, and they're just releasing the new software version for that year. You can either renew for $25, or get the new software for $40 -- wait long enough and rebate drops it to $15.

So what would you expect from a shareware offer?

On my end of things, I went back to the two web-sites -- for the 3.4 version and the M____86+ v.1.70 (which I'm currently using). the ""+" is supposed to be multi-core capable, and the 3.4 version incorporates "1.70" multi-core features. So at least, I have an incentive to try 3.4, and should remember to resurrect this thread if it gets "buried" and let you know. 3.4 page says "recognizes newer chipsets."

However they coded it -- in C or in Assembly, even an Assembly-language program could be composed with more bells and whistles and a little more finesse, without adding much bloat.

 

[BonzaiDuck]

FOR "MadScientist" per my "promise to get back and 'touch base'" ----

OK. Here it is -- an assessment thus far.

Don't want to take your time. You were concerned about the ISO disc creation -- a CD-ROM that will boot into MEMTEST86 3.4.

BEFORE I even TRIED to cut a CD with the program, I made a floppy. I'm in the middle of testing two sets of 2x1GB Crucial Tracer DDR2-800's in a hardware setup I'm planning to use with VISTA 64-bit. The OS has not been installed.

I started running MEMTEST86+ v 1.70, and except for the first attempt to set the system "roughly" to latencies and over-clock settings I'd seen with regular Crucial Ballistix DDR2-1000's, it has so far showed no errors.

While I was running that, I went to a proven P4 system with only a GB of PC1066 32-bit Samsung RAMBUS memory -- reliable, tested several times -- never had a problem with it -- never -- NEVER -- overclocked this latter system.

I used this latter system to create a MEMTEST86 3.4 floppy.

Then, I halted my tests in the system with the Crucial Tracers, rebooted, and watched MEMTEST86 3.4 start up. After it got through test #2, the upper screen went blank and the lower screen displayed something that looked like statistics for errors that had ALREADY BEEN TALLIED for all tests through TEST #10 !!!

I rebooted (reset button), and began the test with MEMTEST86+ v1.70 again. I took the MEMTEST86 3.4 floppy, and put it in the RAMBUS system, rebooted that system, and started up the program.

The same thing happend. The same . . .. g__d__n . . . . thing.

Again, I'd wonder how well the $10 CD runs -- if "suddenly" -- everything is "copacetic." But since I KNOW the older system is tip-top, error-free, I'm pretty sure that the download for v 3.4 is buggy -- and I'm willing to speculate -- just speculate -- that it is deliberately buggy.

You can either stay away from this "revised" MEMTEST86 altogether, or risk spending $10, whereupon (you would pray) that the progrram suddenly works fine and shows no errors on an error-free system, as if "by miracle."

That's about as far as I go with this version of MEMTEST86, given what YOU (MadScientist) KNOW, and what I'VE FOUND OUT.

 

[MadScientist]

Thanks for your efforts Duck. I'm going to contact them in their Problem and Bug Reporting section and let them know what a POS program this is. Link

 

[BonzaiDuck]

Good. Your attitudinal strategy with the custodians of MEMTEST86+ reflects my own in regard to my dead-beat tenant squatting (for truly limited time) in my Virginia property.