Ram stick dead after short usage.

TXAV8R

Commendable
Oct 31, 2016
2
0
1,510
After leaving my pc in sleep mode all night, my computer woke up and gave me the BSOD "Memory_Management" error. After reboot it BSOD'd with the same error right as I entered the desktop. I rebooted again and was able to keep windows running for around 5 minutes and then the same error happened again while opening a tab in Chrome. After another reboot, I started to get some other random BSOD errors that indicated other problems which confused me. I then finally ran windows memory diagnostic. Immediately it detected "hardware issues" during the first pass. I took out one stick of ram and ran the test again and found no hardware errors. Swapped with the other stick and again, immediately a hardware issue was detected, narrowing it down to one stick. After exiting and rebooting on the faulty stick I once again got a BSOD right as windows booted to desktop. I did these tests in the same slot just to see if it's a slot issue, but it doesn't look like it. I haven't run memtest86 and already RMA'd the stick because it was pretty apparent to me what the issue was. My question is, why did the stick fail so fast? I built the pc 48 hours before it happened, so it's brand new. However, I have overclocked my Ryzen 1600 to 3.8 ghz with around 1.35 volts and I DID tamper with SOC voltage to run at around 1.2 volts. I really shouldn't have adjusted that setting, seeing as I don't fully understand what it does (tutorials and guide following is what led me to adjust the SOC voltage). The cpu runs very cool though, at around 55-60 celsius at high load. I also ran my 16 gb of DDR4 ram at the rated speed of 2400 mhz but I didn't mess with timings and voltages and I didn't run the DOCP profile because I heard that it was iffy. The motherboard seemed to do a good job of automatically adjusting these settings but I could be dead wrong and my ram was either being given too little voltage or too much and was overheating (I never monitored the ram during gaming or running applications). Apologies for this mini book of a post but I'm cramming information so that I can get some answers as to what may have caused this so I can avoid it happening again.

PC specs:
MOBO: asus rog strix b350
CPU: ryzen 1600 3.8 ghz
PSU: corsair cw650 80 bronze
Memory: trident z rgb 16 gb ddr4 2400
GPU: gtx 1080


 
Solution
Keep the voltages under the limits set for the memory used. Plus keeping the RAM cool is probably the two things you can do to limit memory problems.

CPU coolers that don't provide air flow to the RAM can cause thermal problems. For example, water blocks cool the CPU but do absolutely nothing to cool the RAM. A RAM cooler is a good idea for those applications.
 

TXAV8R

Commendable
Oct 31, 2016
2
0
1,510


Well I'm not exactly sure what voltages the ram was running at, especially when gaming, so I'll try to keep that in check next time as I took that for granted. As for the CPU, I'm using the stock wraith spire cooler which intakes airflow directly onto the heatsink. My ram modules were also sitting very close to the cooler and the cpu. Maybe in theory this could be blowing hot air around and near the ram modules causing thermal issues, but the cpu was running at very stable temperatures and was never overheating. Also, I don't see the necessity for putting a cooler over the ram because I'm not overclocking at insane speeds but at speeds the ram is rated for (2400 mhz). I'll see what happens with the replacement, I really hope the issue isn't with the motherboard and DIMM slots.
 


The suggestion of a RAM cooler was more for water cooled CPU applications. The heat problem is also for situations where all four slots are populated.
 
Solution