[opensuse] memtest86+ .. Recent bad memory for OCZ? Anyone else? [slightly ot]
Listmates, I have received and had to RMA 2 sets of OCZ Platinum DDR2 1066 in the past week (the first set painted the memtest86+ screen red right off the bat, the second set (2 2-gig sticks), just one stick was bad that failed test 7). Literally, the replacement RMA set was bad. What gives? Bad run of silicon? Anyone else bought any in the past few weeks? Surprising, because other other than a couple of prior OCZ RMAs over the years, I really haven't had any problems with OCZ. For the on-topic part. If anyone is relatively new to memory errors, then you need to know memtest86+ is your friend. If you experience stray lockups or other random hardware failures (mouse just quits, usb devices (drives) just disappear or quit working, etc..), then install memtest86+ which becomes a grub menu boot option you can select to boot into memtest. Generally, you just start it and go do something else for a while and let it run (it will loop through its suite of tests continually and each complete pass can take up to an hour or more depending on how much memory and your other hardware specifics) You can also configure memtest86+ to some extent while it is running. Press 'c' and you can skip tests, reset, etc.. How many memory errors are acceptable - ZERO! If you see any red on the screen you have problems. Do all problems mean fault memory - NO. You need check your bios to make sure your bios memory setting (CAS timings, etc.) are not set to 'tight' or 'lower' than what your memory specifications are. Try loosening/raising the bios memory timings or set them to "Auto" and retest before you throw in the towel. If your settings are set to "Auto", which most are, then you should be fine on the settings. If after confirming your bios settings, you still see red in memtest86+, it's time to RMA or warranty your memory. You can still narrow down which stick it is by removing the sticks and testing them one-by-one, but if you purchased a dual-channel kit, they're both going back anyway... Now back to my question, anybody else seeing bad sticks of OCZ DDR2 lately? Thanks. -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Another thing to check is the voltage. I have some OCZ RAM in my system and the voltage is not the 1.8v DDR2 standard but instead it is 2.1v. The BIOS did not pick this up automatically. It would run ok as 1.8v but when I added a 2nd pair of RAM it would behave oddly, bumping it up to the correct voltage fixed the problem. Some systems won't let you adjust these "overclocking" settings. And of course, if you have multiple pairs of bad RAM I would be inclined to check the mainboard and CPU (recent AMDs have the memory controller on the CPU -- I walked barefoot accross a carpet with a CPU in my hand and it never worked in dual-channel mode again!) -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Thursday 11 June 2009 08:11:30 pm Andrew Joakimsen wrote:
Another thing to check is the voltage. I have some OCZ RAM in my system and the voltage is not the 1.8v DDR2 standard but instead it is 2.1v. The BIOS did not pick this up automatically. It would run ok as 1.8v but when I added a 2nd pair of RAM it would behave oddly, bumping it up to the correct voltage fixed the problem. Some systems won't let you adjust these "overclocking" settings.
And of course, if you have multiple pairs of bad RAM I would be inclined to check the mainboard and CPU (recent AMDs have the memory controller on the CPU -- I walked barefoot accross a carpet with a CPU in my hand and it never worked in dual-channel mode again!)
Thanks, This was confirmed bad RAM. I independently tested the sticks, and then on the 3rd RMA sent it to OCZ under warranty. It was bad. OCZ was great though, and pulled the return and tested it before sending the replacements back. I don't know how it gets packaged, or whether NewEgg had something to do with the problem, but there was one good stick and one bad stick in each of the *three* RAM kits NewEgg sent me. -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Thursday 11 June 2009 08:11:30 pm Andrew Joakimsen wrote:
And of course, if you have multiple pairs of bad RAM I would be inclined to check the mainboard and CPU (recent AMDs have the memory controller on the CPU -- I walked barefoot accross a carpet with a CPU in my hand and it never worked in dual-channel mode again!)
I must be lucky messing with processors. I am careful with all components, but that is usually limited to grounding myself by touching the bare metal of the power supply while the power cord is still attached, removing power, give a 30 sec. bleed down for residual power, then just rip the processor out, spray it down with WD-40 to remove all the old thermal paste, etc, clean up the excess WD-40 with a paper towel, hit it with a small blast of WD again, then take a Q-tip for final clean up of the chip and then use a soft cotton cloth to eliminate any dust/lent (an old t-shirt works fine, or old "cotton flour sack" dish towel) and then reinstall. Probably 20 or so processor replacements and they all still work (knocking loudly on wood ;-) -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
David C. Rankin wrote:
Listmates,
I have received and had to RMA 2 sets of OCZ Platinum DDR2 1066 in the past week (the first set painted the memtest86+ screen red right off the bat, the second set (2 2-gig sticks), just one stick was bad that failed test 7). Literally, the replacement RMA set was bad. What gives? Bad run of silicon? Anyone else bought any in the past few weeks? Surprising, because other other than a couple of prior OCZ RMAs over the years, I really haven't had any problems with OCZ.
I haven't bought any of that memory, but a couple of years I bought some for my ThinkPad. At first I tried an off brand and both original and replacements failed with memtest. I then bough Kingston and it has worked fine and no memtest errors. -- Use OpenOffice.org http://www.openoffice.org -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 2009/06/11 18:07 (GMT-0500) David C. Rankin composed:
Now back to my question, anybody else seeing bad sticks of OCZ DDR2 lately? Thanks.
I have not bought any 1066, and no OCZ lately (only once actually, PC5300 quite a few years ago). Last I bought was Kingston ValueRAM PC6400 2x1G. It was/is fine. Before that last was Crucial Ballistix Tracer PC6400 2x2G. One stick had one randomly bad bit that was not helped by increasing voltage. The warranty replacement pair have been flawless at stock voltage. OTOH, I agree with Andrew that voltage can be the problem, but disagree with his description. AFAIU, a RAM stick's voltage spec is a voltage limit, that is, the spec is the maximum RAM voltage tolerated by the part, which means higher spec is at least theoretically better for overclocking. In my experience over the years, if you have one or a few bit errors, a voltage increase may make them go away. More likely if the supplied voltage is too low, the stick won't be recognized at all, and may prevent POST. -- "Cast but a glance at riches, and they are gone, for they will surely sprout wings and fly off to the sky like an eagle." Proverbs 23:5 NIV Team OS/2 ** Reg. Linux User #211409 Felix Miata *** http://fm.no-ip.com/ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
participants (4)
-
Andrew Joakimsen -
David C. Rankin -
Felix Miata -
James Knott