X99 experienced a loss in a memory channel.
X99 experienced a loss in a memory channel.
Default bios settings look reasonable now. Auto RAM is using JEDEC 2133 timings. It’s possible some configurations were off even though the system started without issues after swapping the CPU.
I've seen comparisons for X99 as well, but nothing quite like this before. Still, if I had to point it out, it's likely the mobile BIOS. I couldn't locate an XMP toggle option, and it's present in the newest BIOS versions. The default loading made it function as intended, so I suspect some configurations weren't fully restored when I swapped the CPU. It seemed to work enough initially, but a proper reset was needed for accurate values. Others suggest resetting settings after a BIOS update to avoid issues, and I've noticed some odd behavior if you don't do that. Depending on the motherboard, many updates may require a reset anyway, but it's worth checking. In the end, it's running and it doesn't stress me with 12GB RAM.
I've heard various opinions about it, some blaming the mounting pressure. For years, I had issues on my X58—it was quite finicky. But! The board features two BIOS versions; one had the latest, the other an older model. Switching to the older one cleared up all my memory problems. It took me seven years, but I eventually forgot it existed. I went from needing 6GB RAM to being able to fully use all dimmer settings with three different kits. I could even run more than one type at once. For example, my main RAM was STT Chrome Hypers, but running G.Skill BBSE would cause issues for a while. Especially when trying to use my OCZ PSC. With the older BIOS, all those kits worked just fine along with 8GB of G.Skill Hynix cards. That's pretty wild. And honestly, sometimes I'd just drop channels—they'd appear and disappear instantly.
The board is the main component, not the CPU or RAM. I've encountered this with many AsRock boards, including one that performed similarly. It was a very powerful 4-core processor.
I attempted a fix by adjusting the screws that secure the waterblock. The main issue only occurred with another AIO model using 115x, where temperatures spiked under tighter settings. That didn’t affect me this time. I switched to Asrock, mainly because of its performance rather than other factors. I also have Z370 and B450 boards; the power delivery seems insufficient for stable operation during long tasks. The Z370 tends to crash intermittently unless turbo is disabled, after which it runs smoothly (4.3 GHz vs 4.0 GHz all cores). The B450 fails differently based on CPU models, worsening with newer releases. Limiting power didn’t help and sometimes made things worse. Only a stable voltage offset improved things. There are affordable OC boards from reputable makers, but I’m cautious about MSI and Asrock. Their BIOS layouts are poor, updates lag, and RAM compatibility is an issue with Z170 boards versus Asus. Asus is still a contender, though the last unit I got was X299. I’m considering Gigabyte’s Z490 now, especially since I prefer early purchases for BIOS updates.
Just a note, I faced a very similar problem. My ASRock X99m Fatality had a 5960X before, and one memory channel failed on that board. After trying almost everything and giving up, I kept using it with 12GB (4x4GB at the time). Interestingly, it worked again on its own. Later, after a month or two, it stopped functioning and eventually died. The CPU was fine, as I tested it on an EVGA board, so I think the issue might be with the socket or traces. Since the build was quite old—5 to 6 years—I disassembled everything and sold the parts. The lesson? No more ASRock boards for me.
They produce a solid board, though it's quite small. It resembles the BioStop boards in that regard.