Hi,
I'm having a weird issue with the latest Memtest V5.0b8 on a MacBookPro i7-2675QM 2.20GHz 2011, with DDR3 1600MHz PC3-12800:
* Block move 64-byte block is the only test to fail (14 to 18 errors with 16GiB)
* All other tests pass w/o any error with the default Memtest settings.
1. The test passes Ok with the original DDR3 memory (2x 2GiB)
2. The test fails with 16 errors with the new DDR3 memory (2x 8GiB)
I first suspected one of the two new memory devices to be in trouble. However:
1. The test passes Ok with a single 8GiB memory device (1x 8GiB). I tested each memory device alone: no error. I also switched the memory slots (upper, lower) with each device: no error.
2. The test passed Ok with a mix of old and new memory (1x 2GiB + 1x 8GiB). I tried various combination of memory device and memory slots: no error.
In other words, the test only fails with both 8GiB devices are used. Unfortunately, I do not have a spare memory device of this kind to perform a test with another 8GiB memory device.
The test is always failing with the same error bit mask
Bits in Error Mask 0000000080000000
However the actual memory addresses in error seem more variable (with the same HW/memory config). Two subsequent test do not report the same in-error addresses, nor the same error count (14 to 1.
Addresses in error seem to cover a wide range:
38ace20d0
388ce20f0
33b842d10
339842d30
14a0ca3d0
1480c3f0
43f3f3e10
43d3f3e30
5be29110
I'm not sure how to interpret these results: is there an actual HW issue, or is there an issue with the test?
BTW:
* would there be a way to store the faulty addresses with the generated output report?
* is there a way to create a (text) snapshot of the displayed info?
Thanks,
Emmanuel.
I'm having a weird issue with the latest Memtest V5.0b8 on a MacBookPro i7-2675QM 2.20GHz 2011, with DDR3 1600MHz PC3-12800:
* Block move 64-byte block is the only test to fail (14 to 18 errors with 16GiB)
* All other tests pass w/o any error with the default Memtest settings.
1. The test passes Ok with the original DDR3 memory (2x 2GiB)
2. The test fails with 16 errors with the new DDR3 memory (2x 8GiB)
I first suspected one of the two new memory devices to be in trouble. However:
1. The test passes Ok with a single 8GiB memory device (1x 8GiB). I tested each memory device alone: no error. I also switched the memory slots (upper, lower) with each device: no error.
2. The test passed Ok with a mix of old and new memory (1x 2GiB + 1x 8GiB). I tried various combination of memory device and memory slots: no error.
In other words, the test only fails with both 8GiB devices are used. Unfortunately, I do not have a spare memory device of this kind to perform a test with another 8GiB memory device.
The test is always failing with the same error bit mask
Bits in Error Mask 0000000080000000
However the actual memory addresses in error seem more variable (with the same HW/memory config). Two subsequent test do not report the same in-error addresses, nor the same error count (14 to 1.
Addresses in error seem to cover a wide range:
38ace20d0
388ce20f0
33b842d10
339842d30
14a0ca3d0
1480c3f0
43f3f3e10
43d3f3e30
5be29110
I'm not sure how to interpret these results: is there an actual HW issue, or is there an issue with the test?
BTW:
* would there be a way to store the faulty addresses with the generated output report?
* is there a way to create a (text) snapshot of the displayed info?
Thanks,
Emmanuel.
Comment