Is my memory f*cked?

[diab10]

Memtest 3.2 and Memtes86+ V1.6 reports two erors at this adress: 000097557d8 - 151.3 MB adresa. Should I disregard this error or should I ask my saler for a new memory (single module 512 Kingmax)

 

[bob4432]

get a new stick. you will probably error out in prime95 too, an error is an error. let memtest run overnight and see how many you get.

an error is an error 🙁

 

[SrGuapo]

How many passes did you run? Anmy amount of errors is bad. Are you OCing or doing anything out of the ordinary with the RAM (changing the voltages, etc.). If this is a new stick, contact Kingmax/retailer and request an RMA.

 

[diab10]

well i ran it on v3.2 for an hour an it reported these two errors, on v1.6+ i ran it for 8 hours, and it still reported "only" two errors. no overclocking or any voltage changing. MB Epox 8kda3j, this board has a little voltage variables issues, but it isnt that bad.

 

[diab10]

i had BSOD, and problem reinstaling XP, safe mod stoped at mup.sys, and only after total deleting of partitions and formating with seagate disk wizard (i have 120 gb Seagate sata) i installed win, but the question is for how long with this memory stick.

 

[bob4432]

Originally posted by: diab10
well i ran it on v3.2 for an hour an it reported these two errors, on v1.6+ i ran it for 8 hours, and it still reported "only" two errors. no overclocking or any voltage changing. MB Epox 8kda3j, this board has a little voltage variables issues, but it isnt that bad.

what kind of memory - pc3200, pc133? if it is pc3200 3.2V seems pretty high as i think most is ~2.6V

 

[SrGuapo]

A couple errors shouldn't bother you too much in Windows, but you still should get a new stick. Did you just buy this RAM, or is it fairly old? Any errors are bad and shouldn't occur on healthy RAM.

 

[diab10]

Its pc3200, an the modul is brand new, so i think i`m gonna return it, thay have to acknowledge memtest errors dont they? 😉

 

[bob4432]

Originally posted by: diab10
Its pc3200, an the modul is brand new, so i think i`m gonna return it, thay have to acknowledge memtest errors dont they? 😉

in theory yes, but 3.2V seems pretty high...do you have active cooling on it? what fsb did you get to? i wouldn't tell them you went that high unless it is supposed to support that V.

 

[diab10]

ups sorry you missunderstood me i talked about version of memtest i was using v3.2 and v1.6+, thanx anyway

 

[bob4432]

Originally posted by: diab10
ups sorry you missunderstood me i talked about version of memtest i was using v3.2 and v1.6+, thanx anyway

oh, ok, that sounds better, just didn't want you to tell the rma peeps you were running it @ 3.2V:Q

 

[RobCur]

Originally posted by: diab10
well i ran it on v3.2 for an hour an it reported these two errors, on v1.6+ i ran it for 8 hours, and it still reported "only" two errors. no overclocking or any voltage changing. MB Epox 8kda3j, this board has a little voltage variables issues, but it isnt that bad.

I never ran it for an hour let alone half an hour. did you computer crashes? is your computer sluggish often? then you have nothing to worry about. these errors are normal and are correctable by your cpu. no memory module are perfect, their will always be errors at some degree and a little is tolerable.

 

[bob4432]

Originally posted by: RobCur

Originally posted by: diab10
well i ran it on v3.2 for an hour an it reported these two errors, on v1.6+ i ran it for 8 hours, and it still reported "only" two errors. no overclocking or any voltage changing. MB Epox 8kda3j, this board has a little voltage variables issues, but it isnt that bad.

I never ran it for an hour let alone half an hour. did you computer crashes? is your computer sluggish often? then you have nothing to worry about. these errors are normal and are correctable by your cpu. no memory module are perfect, their will always be errors at some degree and a little is tolerable.

i beg to differ, i can run my machines for 24+hours with no errors in memtest or prime95...

 

[RobCur]

then again these errors are caused by your cpu miscalculation. hehe and it has to correct itself when it make mistakes. their is a thing called ecc error enabled on most motherboards in the bios, which mean that the memory module don't cause errors if you're using software memorytester, these are next to useless imho and time consuming.
if windows use swap file of 700mb and give you 0mb often in taskbar where your hd is constantly accessing then you have motherboard problems or is compatible memory module with your mobo
dimm either work or they do not.

 

[bob4432]

Originally posted by: RobCur
then again these errors are caused by your cpu miscalculation. hehe and it has to correct itself when it make mistakes. their is a thing called ecc error enabled on most motherboards in the bios, which mean that the memory module don't cause errors if you're using software memorytester, these are next to useless imho and time consuming.
if windows use swap file of 700mb and give you 0mb often in taskbar where your hd is constantly accessing then you have motherboard problems or is compatible memory module with your mobo
dimm either work or they do not.

where is the ecc error item in the bios? isn't that only if you are using ecc ram, like a server type board? 0MB in the taskbar - what is this?

 

[CrispyFried]

ECC memory is special error correcting memory (you usually know if you have it, its a selling point) that can correct certain memory errors (not all though). If you have ECC and its enabled memtest WILL NOT see those errors, its all corrected in hardware before the CPU or anything else sees it.

Most memory in consumer rigs is not ECC, ie any errors are treated as real data, with predictable results (BSODs, file corruption etc).

Any error with memtest is bad. Ive run memtest for 24+ hours with no errors.