XMP Profile 1 experiences issues on the Kingston Fury Renegade 7200.
XMP Profile 1 experiences issues on the Kingston Fury Renegade 7200.
Hello, I’ve set up the following configuration: ASUS ROG MAXIMUS HERO Z790 Intel i9-13900K with 32GB (2x16) Kingston Fury Renegade at 7200MHz, XMP I and Profile 1 enabled at that speed. MCE is turned off, CPU is capped at 253W. I’m experiencing frequent crashes—Steam, Office apps, Chrome STATUS_ACCES_VIOLATION, Windows Explorer crashes, and game crashes after some playtime. During the AIDA64 stability test, hardware failure occurs after 30 seconds and BSODs appear with error code MEMORY_MANAGEMENT. Despite running Windows Memory Diagnostic or MemTest86 multiple times, I still see no errors. When I lower the frequency to 6933MHz while keeping timings and voltages constant, there’s a noticeable improvement—AIDA64 no longer reports hardware issues, crashes become less frequent, though they still happen. The OCCT memory test runs smoothly for 15 minutes at that speed, but it wasn’t run longer. It seems the instability sometimes depends on luck; I’ve had days with zero errors in the reliability monitor, only to see Explorer or Steam crash immediately after startup and need a restart. For a few weeks now, I’m trying to figure out why my system is unstable. I don’t believe there’s a CPU or memory problem, but think it might be related to BIOS settings or driver issues—ASUS ARMOURY CRATE, NVIDA GeForce Experience, and Windows Update are installing drivers. Thank you!
Having worked extensively with DDR5 overclocking recently, achieving speeds above ~7000MT/s remains quite challenging for stability. Kits targeting 7200MT/s offer solid performance, but they demand significant effort to maintain consistency. Simply enabling XMP isn’t sufficient for reliable results. The most common problem is instability in the memory controller. A useful stress test to detect these issues is Y-Cruncher’s VST test; frequent blue screens during testing often indicate a crash within the first run with your current configuration. If you aim for stability at 7200MT/s, focus on the VST test. Three voltage settings—VDDQ TX, VDD2, and VCCSA—most influence controller stability. On ASUS boards, these may be labeled differently (ASUS tends to use alternate names), so check for IVR Transmitter voltage, Memory Controller voltage, and System Agent voltage. These three values usually align in a stable range, though the optimal setting shifts based on BIOS version, memory speed, and their interaction. Adjust each until IVR Transmitter stays under 1.45V and System Agent below 1.45V (ideally even lower), while keeping Memory Controller voltage near 1.4V. Aim for below 1.8V on the controller voltage—though this is more about guidance than a strict target. If you’re unfamiliar with ASUS hardware, I can’t confirm exact preferences, but from screenshots I’ve noticed around 1.4V for IVR Transmitter and 1.3V for Memory Controller with 1.45V for VCCSA as possible benchmarks. This gives a starting point, though results vary by board model. Alternatively, consider using a kit rated for 6400MT/s; it’s a practical choice since 7200MT/s is achievable and still requires manual voltage tuning. Hynix A-die kits are involved here, so they’re likely to overclock successfully if you invest the time in proper settings. The main trade-off is cost versus performance—going with a higher-rated kit saves money and ensures XMP functionality, while tuning later can still unlock 7200MT/s.
Thank you for your feedback. I'm considering switching to a different kit but need to complete this project right now. I attempted the second XMP profile at 6800MHz with altered timings and voltages, which caused a BSOD after 10 seconds. I returned to the first profile at 7200MHz, maintaining original settings but lowering the frequency to 6800MHz. It seems stable so far, though I haven't tested the Y-Cruncher VST yet. The CPU is also undervolted by 95mV—could that be affecting performance?
Technically feasible, yet this only makes sense if your undervolt wasn't actually stable from the start. I’ll keep pointing out that your CPU’s memory controller voltages aren’t properly configured by the BIOS using auto settings, which is likely the root cause. It might also be due to a CPU with a very limited IMC—7200MT/s is near its maximum—and pushing chips to their limits requires extensive tuning, or the chip has unusual voltage needs.
The crashes occurred before the undervolt event. Once I lowered the CPU voltage, I ran 30-minute Cinebench R23 tests (results around 39500-40500) and AIDA64 stability checks (15-30 minutes), no BSODs, no hardware issues. All runs used XMP Profile 1 but reduced frequency from 7200MHz to 6933MHz. The system felt stable for a while then crashes started, eventually leading to BSODs—randomly, regardless of load. Recent error codes included CRITICAL_PROCESS_DIED (svchost.exe), IRQL_NOT_LESS_OR_EQUAL, and SYSTEM_SERVICE_EXCEPTION. Could this be related to drivers? I considered reinstalling Windows and fixing drivers manually. It’s confusing since memory tests showed no errors with XMP enabled, and benchmarks/stability tests didn’t cause crashes.