Why does Memory Fail?

[johnbullas]

XP generic system (Neo Platinum Something MB with Athlon 64 3200+) with 1 x 1gb and 1 x 512Mb DDR400 184 pin ram wafers

XP started hanging today (locking up not blue screening) and after doing a system restore (to before a recent driver update) it still left XP hanging after a random few minutes

I thus resorted to running the memtest86 boot disk

http://www.memtest86.com/

the auto test on boot promptly HUNG after both of 2 reboots but did not hang after the 512MB wafer was removed... it is generic ram marked CM400/512/oem with
1920105 00000891 on the bar code

trying using uptime.exe

http://blogs.msdn.com/gusperez/articles/91734.aspx

to show how long my pc has been running, guess the frozen clock in the tray will tell me when it stopped if it does

Hopefully the problem is gone along with $60+ of RAM

What actually makes RAM "fail"? can the offending chip be isolated? it was at least 18 months old and not over clocked, machine has been on for a bout 3/4 of that period mind you...

Will try to replace the wafer in the slot but it has NOT been removed or distrubed since it was installed

John
Southampton


=========================
Transport Solutions
Southampton UK

 

[edfair]

There are all sorts of failures. Do a google search on the test names in the memtest series and you should find some descriptions.
Some failures can be diagnosed to specific chips and those could be replaced with the right equipment. Others are ramdom with no way of understanding what happened.
There are timing issues that start coming into play as chips age and if they were designed on the edge of the specs it is possible that the failures are age related.

Ed Fair
Give the wrong symptoms, get the wrong solutions.

 

[G0AOZ]

Make sure the memory strips are not getting too hot. This can be a major cause of failure. Problem can sometimes occur because of poor motherboard design in the placement of the RAM sockets. This coupled with high heat output from say a close spaced CPU cooler, hard drives getting very hot, or badly laid out low voltage wires and data cabling, can conspire to overheat the RAM. Important thing is to carefully check airflow over the memory area.

An overheated CPU will either lock up, or at worst totally fail. RAM chips contain semiconductors, albeit not as many as the CPU, but nevertheless they won't operate at extremely high temperatures. Fatigue or total failure can result. This can also happen if the motherboard is hit by a large voltage spike for instance. Many devices on the motherboard will have in-built protection circuits, but there is only so much electrical "abuse" they can take before they fail.

ROGER - G0AOZ.

 

[SkipCox]

Heat, static electricity, degraded connection with the motherboard, poor voltage regulation, and no apparent reason whatsoever in no particular order.

You can usually use memtest to id the offending chip...for example, a rig with two 256Mb chips that fails at the 370Mb point would point you to the memory in the 2nd slot.

Usually...sometimes...on occasion...

My last go round in a similar situation pointed toward a bad memory chip and a bad motherboard memory slot. I replaced the offending chip and the machine has been just fine ever since. The "offending" memory chip has been working just fine in another machine for over a year.

Go figgur...

Skip

http://skipcox.net

 

[johnbullas]

There have been numerous reports of errors with only tests 5 and 8 on Athlon systems. 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 Athlon speeds. 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 with an Athlon! On occasion test 5/8 errors will occur even with name brand memory and a quality motherboard. These errors are legitimate and should be corrected."

Test 5 [Block move, 64 moves]

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 a 8mb segment of memory the failing address will always be less than 8mb away from the reported address. Errors from this test are not used to calculate BadRAM patterns.

Test 8 [Modulo 20, ones&zeros]

Using the Modulo-X algorithm should uncover errors that are not detected by moving inversions due to cache and buffering interference with the the algorithm. As with test one only ones and zeros are used for data patterns.

THUS when I get the Pentium 4 2.7 build on ther desk next to me sorted out, in goes this wafer!!!!

FB


=========================
Transport Solutions
Southampton UK

 

[cdogg]

John,
Just an FYI that the Athlon 64/X2 CPU uses the fast Hypertransport bus that operates at 800MHz or faster. Therefore, it is HIGHLY recommended that you enable dual-channel on the motherboard when possible.

In this case, you were running a 1GB stick of DDR with a 512MB stick of DDR. This would have disabled dual-channel, which requires matching amounts on each channel. See your motherboard's manual for different memory configurations that will work. Typically, it is also best to use a pair of RAM modules that have the same timing specs for the optimized performance. Buying the same brand and model for each is the easiest way to go.

As for failing memory...

The others have pretty much listed the most common reasons. I would only add that poor handling passing static on contact can also do damage (though that doesn't apply here).

~cdogg
"Insanity: doing the same thing over and over again and expecting different results." - Albert Einstein
[tab][navy]For general rules and guidelines to get better answers, click here:[/navy] faq219-2884

 

[johnbullas]

FYI back atcha ...... Motherboard is an Msi K8N Neo MS-7030 (v1.X) ATX Mainboard with only single memory channel support, Processor is AMD Athlon 64 Newcastle. Will run up the memory on Memtest86 in the P4 machine I'm building.




=========================
Transport Solutions
Southampton UK

 

[cdogg]

LOL, sorry to hear that John! I should have mentioned that the one exception is the early Athlon 64 chipsets that were only single-channel! Figured there was a small chance you'd have one, but I guess you were unfortunate enough to qualify!!


~cdogg
"Insanity: doing the same thing over and over again and expecting different results." - Albert Einstein
[tab][navy]For general rules and guidelines to get better answers, click here:[/navy] faq219-2884