SNB, Z68 & MEMTEST86 v.4.0a vs MEMTEST86+ v.4.2x

[BonzaiDuck]

Has anyone with a Z68 mobo and i7-2600k or i5-2500k obtained strange results, infinite loops or system "freeze" with the PLUS ("+") version?

I have G.SKILL RipJaws "-GBRL" RAM, which tested fine on an earlier version MEMTEST86 or "+", but did not report cache and other information. I had set the RAM voltage at 1.51V; had installed OS, OC'd cautiously to 4.6 Ghz; tested stability for 8+ hours on PRIME95 and through 35 iterations of IntelBurnTest at "Very High" stress.

I wanted to attempt changing the command-rate to CMD=1. Running MEMTEST86+ 4.2, it continued looping through the same test indefinitely either before or after the 2nd "pass." That is, it just kept running Test #8 perpetually, without moving on to the next pass.

So I reset the command-rate to CMD=2. Decided later to run the MemTest "Plus" version again. This time, the system just froze after starting the second pass.

Downloaded the Brady version MEMTEST86 v.4.0a. [the "non-Plus" version and basis for the "plus" version]. Doesn't present as much info as the "plus" version, but . . . no errors.

Who . . . or what . . . can you trust these days?

 

[BTRY B 529th FA BN]

Your RAM/Memory is probably failing aka degrading

 

[krose]

Memtest86+ 4.2 works with my Z68 and 2600k, no lockups, but I do get Blck reported as 79, which gives Ram reported as 1264 instead of 1600. So something isn't right with it.

 

[MadScientist]

Works for me too, but it's reporting my ram running at 1147 instead of 1600, BCLK of 71. Strange.

Try running HCI memtest, see the sticky: http://forums.anandtech.com/showthread.php?t=2195063

 

[Idontcare]

Bonzai, SB here on Z68 and GSkill ripjaw ram as well...my rig has not compatibility issues with memtest+ BUT it will pass memtest+ with flying colors (>24hrs of continuous testing) while failing to pass HCI memtest.

Wherever the instability now lies in these high-speed DDR3 systems, it would appear that the testing algorithm that memtest+ uses is insufficient to test out the corner-case conditions that are now at fault for memory instability with these systems.

I came to appreciate HCI memtest (props to jvroig for hooking me up with it) in a trial-by-fire fashion. My rig would pass memtest+ at stock clocks (stock SB and stock DDR3-2133 for the ram) but would fail IBT.

Underclocking my ram to DDR3-1866 resolved the IBT fails. Then using HCI I was able to isolate my ram as being the culprit...but for whatever reason memtest+ never detected the issues.

 

[BTRY B 529th FA BN]

I've been saying Memtest HCI all this time!

:hmm:

 

[Idontcare]

I've been saying Memtest HCI all this time!

:hmm:

Sorry, all inputs provided by Vitamin D3 junkies are automatically discounted and dismissed with extreme prejudice :sneaky:

 

[BTRY B 529th FA BN]

Ha!