mjs90201

Since installing the new RDRAM, I have only gotten memory-related error messages after running the Norton Diagnostic program. I believe the Norton program is the cause of these errors. My computer still crashes for other reasons (my video adapter/driver is a prime suspect), but that was happening before the memory upgrade. My BIOS does not offer an option to select PC600 or PC800. [i think I need to update the BIOS though.] As far as reinstalling Win ME, I think a better idea might be to upgrade to Win XP. But that's a project for another day. Thanks for the help! -MJS

The RDRAM are identical and are properly installed. Your other recommendations are a bit beyond my understanding at this time. The RDRAM passed all the tests of the memtest86 and windows memory diagnostic programs. Is this not convincing evidence that the RDRAM is good? -MJS

Do you know the false-negative error rate (i.e., failure to find errors when the RAM is in fact defective) of the Microsoft memory test or memtest86? I have already run the memtest86 program and it found no memory errors. Thus, although I do not know the false-negative error rates of these tests, I am close to concluding that my RDRAM is fine and the Norton diagnostic program is defective. Regards, MJS

I have a couple of additional details to add to the puzzle. I now believe that the only time I receive the insufficient memory error message (when I am in fact using only a small percentage of available RAM as verified by a small memory usage guage program) is AFTER I run the Norton Diagnostic memory test (which my RDRAM always fails). The insufficient error message appears after I close the Norton program and attempt to open another program. After I close the frozen program that I attempted to open, a new entry appears in the cntrl-alt-del program list: "route" What is the significance of the "route" program? -MJS

Until recently, I was running my Dell Dimension 8100, P4, Windows ME system with 128mb of PC600 non-ECC RDRAM. Although my system had frequent crashes which sometimes seemed related to memory limitations, I never had reason to believe the RDRAM was defective. I suspect most of the crashes have been due to compatibility issues with my ATI Radeon graphics card/drivers and/or the drivers for my mouse and keyboard; my virus protection is up to date and virus scans have shown no infections. I recently upgraded my memory to 512mb of PC800 ECC RDRAM (2x256mb). My computer recognized the change in system memory, and now lists 512.0mb RAM in "system properties" instead of 128mb. However, my system crashes and memory-related error messages have increased. Sometimes after I start my computer with a minimal number of start-up programs and attempt to open my first program, I get an insufficient memory error message! This never happened before. Although I no longer use the Norton Anti-Virus program, I still have Norton System Works (2001) installed and I ran their "Norton Diagnostics" program to test the new memory. The first test or pattern which this program runs is described as: "Testing memory using pattern with all bits cleared..." Each time I run this program, after about one second it indicates "Error detected at address xxxxx" and "Test: Failed." Every time I redo the test it indicates an error at the same address. After I reboot, however, the address of the error changes (and remains consistent until I reboot again). In an attempt to confirm that my new memory was defective, I researched other memory testing programs and discovered Microsoft's "Windows memory diagnostic" program, which runs a "standard" set of 6 memory tests and an "extended" set of 11 memory tests (from a bootable floppy disk). My new memory passed the standard set of 6 tests so I switched to the "extended" test set and let it run overnight. In the morning, I discovered that my new memory passed all the tests with no errors. Thus, it seems to me that either the "Norton Diagnostic" program is defective or my new RDRAM is defective. Which is more likely? Does the Norton Diagnostic program use a test (i.e., using a pattern with all bits cleared) which can detect errors in ECC RDRAM that the Microsoft diagnostic program cannot detect? Is there another explanation? I would appreciate any helpful suggestions! Regards, MJS