>Hi, > >I was intrigued by your mention of the "RAM sandwich" method >of testing memory, since my J700 running OS X is plagues with >system freezes. But when I searched the archives under >"sandwich", I only got 4 hits (I asked to display 1000000 items), >only two of which referred to RAM testing, one of which was >your post that I'm replying to now. Perhaps the testing method >was initially referred to by another name? If you could help me >find the info in the archives, I would appreciate it very much. > >Thanks, > >Ken
Ken, I've kept a fairly sizable archive of my own and searched it. There were two helpful posts about RAM sandwiches in April, 2002. From Nancy: "I have a couple of questions about RAM sandwiches and RAMometer. RAMometer came as a freebie with the c600. However, I have never used it. Using the program is pretty straight forward. My question is about error messages. Is an error message an automatic indication that the RAM chip is faulty? Or, is there a margin of error in the error messages? I could not find any documentation on RAMometer. Does anyone know what the error messages indicate/mean? I got a couple of messages and they are pretty much in code. (They give an address, and what the test was shooting for and what it actually got) I guess I need a secret decoder ring! Finally, is my memory correct that the RAM sandwich test is: One new/unknown stick sandwiched between two "known to be good" sticks, and run RAMometer for 500 passes. Is that correct... or was that 5000 passes?" From Jeff: "No errors are acceptable. A computer program may run through a loop millions of times in just a few seconds. If the data it gets isn't accurate every time the program isn't going to run correctly. Nancy, your earlier description of the RAM Sandwich is accurate, with one addition. All three (or four) DIMMs need to be in a single bank, either bank A or Bank B on the S900/J700 (the C series only has one bank). So what is usually described is to put a good DIMM in A1, another good DIMM in A4 and one or two DIMMs to be tested in A2 and/or A3. You could do the test in bank B instead of A. Experience seems to show that about 1500 iterations is sufficient. I don't think anyone has gotten a failure above 1500 without having a failure sooner. However, some of us have received failures at about 1200 - 1300, so it is important to do that many iterations. It's time consuming, which is why several list members suggest just buying from someone with better quality control. Personally, I think it's a good idea to test your RAM no matter who you buy it from, but it does seem that you're more likely to do the RMA shuffle if you buy from Velocity. Jeff Walther" -- SuperMacs is sponsored by <http://lowendmac.com/> and... Small Dog Electronics http://www.smalldog.com | Refurbished Drives | Service & Replacement Parts [EMAIL PROTECTED] | & CDRWs on Sale! | Support Low End Mac <http://lowendmac.com/lists/support.html> SuperMacs list info: <http://lowendmac.com/supermacs/list.shtml> --> AOL users, remove "mailto:"; Send list messages to: <mailto:[EMAIL PROTECTED]> To unsubscribe, email: <mailto:[EMAIL PROTECTED]> For digest mode, email: <mailto:[EMAIL PROTECTED]> Subscription questions: <mailto:[EMAIL PROTECTED]> Archive: <http://www.mail-archive.com/supermacs%40mail.maclaunch.com/> --------------------------------------------------------------- >The Think Different Store http://www.ThinkDifferentStore.com ---------------------------------------------------------------
