Awhile back I reported random crashes.

Testing memory for hours on-end using the latest version of Memtest86
found nothing.

With some time on my hands this week, I pulled the machine out,
swapped out the second bank of memory... problem solved.

Aaaaargh!

Is there a REAL memory test out there?

...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona Voice480)460-2350 | |
| E-mail Address at Website Fax480)460-2142 | Brass Rat |
| http://www.analog-innovations.com | 1962 |

America: Land of the Free, Because of the Brave

On Thu, 11 Oct 2007 16:05:27 -0700, Jim Thompson
wrote:

Awhile back I reported random crashes.

Testing memory for hours on-end using the latest version of Memtest86
found nothing.

With some time on my hands this week, I pulled the machine out,
swapped out the second bank of memory... problem solved.

Aaaaargh!

Is there a REAL memory test out there?

...Jim Thompson

Isn't that what you just did?


D from BC

Jim Thompson wrote:
Awhile back I reported random crashes.

Testing memory for hours on-end using the latest version of Memtest86
found nothing.

With some time on my hands this week, I pulled the machine out,
swapped out the second bank of memory... problem solved.

Aaaaargh!

Is there a REAL memory test out there?

...Jim Thompson

The flip answer is linux. The real answer is no. Memory
gets accessed in so many different ways that a memory test
that is entirely CPU driven is not going to be able to smoke
out all of the flaws. For instance, your computer's memory
is driven by the CPU, the CPU's cache controller, the video
card's DMA, the network card's DMA, the disk controller's DMA,
.... . If your memory is running a little too slow, or too fast,
one of these memory access methods could exercise the flaw
sometimes, and others may never exercise the flaw.

Additionally, the temperature of the computer box varies with
load, room temperature, disk activity.... The CPU's temperature
varies depending on what you are doing with it (idling, CPU
intensive routines, ...).

So, the only real test for a memory card is to use it heavily
loaded in the system it is going to be deployed in.

-Chuck

On Thu, 11 Oct 2007 16:17:56 -0700, D from BC
wrote:

On Thu, 11 Oct 2007 16:05:27 -0700, Jim Thompson
wrote:

Awhile back I reported random crashes.

Testing memory for hours on-end using the latest version of Memtest86
found nothing.

With some time on my hands this week, I pulled the machine out,
swapped out the second bank of memory... problem solved.

Aaaaargh!

Is there a REAL memory test out there?

...Jim Thompson

Isn't that what you just did?


D from BC

Yep ;-)

...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona Voice480)460-2350 | |
| E-mail Address at Website Fax480)460-2142 | Brass Rat |
| http://www.analog-innovations.com | 1962 |

America: Land of the Free, Because of the Brave

"Jim Thompson" wrote in message
...
Awhile back I reported random crashes.

Testing memory for hours on-end using the latest version of Memtest86
found nothing.

With some time on my hands this week, I pulled the machine out,
swapped out the second bank of memory... problem solved.

Aaaaargh!

Is there a REAL memory test out there?


Not with any software. Take the module(s) to a local PC repair shop. Some do have the sophisticated memory testers.

Cheers

Jim Thompson wrote:
Awhile back I reported random crashes.

Testing memory for hours on-end using the latest version of Memtest86
found nothing.

With some time on my hands this week, I pulled the machine out,
swapped out the second bank of memory... problem solved.

Aaaaargh!

Is there a REAL memory test out there?

No.

Jim Thompson wrote:

Awhile back I reported random crashes.

Testing memory for hours on-end using the latest version of Memtest86
found nothing.

With some time on my hands this week, I pulled the machine out,
swapped out the second bank of memory... problem solved.

Aaaaargh!

Is there a REAL memory test out there?

...Jim Thompson
Memory testing programs don't always test the memory in the
all its failing.
it could be a specific order of bit values and address
location change rate that causes the memory to fail.
I've seen a very unique tester module that have an elevation
board that plugs into your memory socket. you then plug your
original back into this carrier board which has enough memory
already on it to accommodate with what your testing.
You run the PC for a while to wait for a crash, if an error
in memory occurs, an LED on this module lights up with a mini
LCD display giving you the address and byte values of the failure.

Basically, as the bus is addressing your memory via the jumper
card, the onboard memory is also being written to and then compared
with your memory at some point to test for verification of integrity.

this is suppose to be a better test to insure the memory is working
correctly in your MB with the settings you have selected. It does not
mean the memory is permanently bad. it just may not be able to operate
at the speed of your settings.



--
"I'm never wrong, once i thought i was, but was mistaken"
Real Programmers Do things like this.
http://webpages.charter.net/jamie_5

"Jim Thompson" wrote in
message ...
Awhile back I reported random crashes.

Testing memory for hours on-end using the latest version of Memtest86
found nothing.

With some time on my hands this week, I pulled the machine out,
swapped out the second bank of memory... problem solved.

Aaaaargh!

Is there a REAL memory test out there?

Were those panels a matched set Jim? If your motherboard was using
an unmatched set in dual channel mode, that could be a problem.

Jim Thompson wrote:

Awhile back I reported random crashes.

Is your power supply going 'iffy'on you ?

It's a classic symptom of the 'bad caps' problem.

Graham

In message , Jim Thompson
writes
Is there a REAL memory test out there?

...Jim Thompson
Right up to the point where you posted that, I would have recommended
MemTest86+ to you wholeheartedly, I suppose the only 'real' memory test
is a dedicated hardware ram tester but it's far cheaper to swap out for
occasional use. FWIW, I've not found a problem with memory that Memtest
didn't also find but obviously it's not infallible.
--
Clint Sharp

In message , IAmTheSlime
TheSlimeFromYourVideo@oozingacrossyourlivingroomf loor.org writes
Boot a Knoppix live DVD disc or even some distros, like Suse Linux or
Ubuntu have memory test selections right from their boot menu, without
even needing to boot into the OS first.

But that's Memtest....
Also, the MS test feature has many switches that allow one to make it a
more comprehensive test. You didn't just select "GO" did you?

--
Clint Sharp

"IAmTheSlime"
TheSlimeFromYourVideo@oozingacrossyourlivingroomf loor.org wrote in
message

Another way to "test" it is to OC the machine in the BIOS. THEN run
the test(s).

A bad stick will show up pretty quickly.

What's OC?

Mobos should have some memory test hardware. Some like mine even have TDR's
to tell you how long your ethernet cable is, which is useless.

For comparison, here's part of an interview with the devolopers of the Alto
at PARC.


The Analytical Engine, Volume 2, Number 1, July 1994 Page 20

CS: They had this neat software that would run every night,
called DMT. And we had a dedicated Alto called Peeker, remember?
And each machine would be left with a kind of screen saver
running, a square that moved around and checked the memory in
all these different locations. And if it got an error when it
was doing this little test, the Peeker would log it.

HY: It was a memory diagnostic that ran overnight. I'll tell
you this. The first Altos had these Intel.... I didn't think it
was going to make it. The 4K chips when they came out were
really okay, reliable, but the 1K Intel chips -- 1130's --
weren't.

CS: Every morning we would check, and anyone who had a bad
chip, we'd go in. The nice thing about it, so many of these
software people didn't know they had a bad or a flaky memory
chip, and a lot of times we'd say "We have detected a bad chip
in your computer and we'd like to change it." And they'd say
"Oh, wonderful!"

--

Reply in group, but if emailing add another
zero, and remove the last word.

On Mon, 15 Oct 2007 11:21:38 -0400, Tom Del Rosso wrote:
"IAmTheSlime"

Another way to "test" it is to OC the machine in the BIOS. THEN run
the test(s).

A bad stick will show up pretty quickly.

What's OC?

Orange County?
Open Collector?
Occasional Carnality?
Odd Circuit?
Of Course?

;-)

Cheers!
Rich

[Ox Cocks? -- Rich the Newsgroup Wacko]

"Rich Grise" wrote in message
news
On Mon, 15 Oct 2007 11:21:38 -0400, Tom Del Rosso wrote:
What's OC?

Orange County?
Open Collector?
Occasional Carnality?
Odd Circuit?
Of Course?

;-)

Cheers!
Rich

[Ox Cocks? -- Rich the Newsgroup Wacko]

Oh, over-clock.


--

Reply in group, but if emailing add another
zero, and remove the last word.