No output / Increase fan speed on GPU / System failure
No output / Increase fan speed on GPU / System failure
The 12VHPWR 12+4 connector always includes 12 power pins plus 4 sense pins. The additional wires you might see missing are merely optional for signaling the GPU about available power limits. They don’t conduct current, so omitting them won’t trigger crashes on their own. Instability often arises from: inadequate connector contact causing overheating, using an adapter or cable unsuitable for your GPU’s power needs, or a faulty EPS or PCIe cable during installation. With the Corsair RM1000e, stable performance under OCCT and gaming points to Loki power supply or cables being the issue. Loki units support all sense wires, but subpar crimp quality could still cause problems. Running stress tests with Corsair confirms the PSU/cables were likely the problem, not RAM, paste, or temperatures. The extra pins are usually harmless distractions.
I agree, I plan to stress test both the new Asus Loki and this Corsair unit. It seems the faulty cables are likely the cause of the issue. I’ll try to confirm it by testing the older Loki as well, hoping to reproduce or trigger a crash. The connectors worked perfectly with the Asus version, even better than Corsair’s, but we’ll find out. I’ll share more details later today or tomorrow, if possible.
Clear strategy for testing both PSUs will speed up the decision. Important point: with 12VHPWR, it’s not just about having all pins, but also depth and contact quality. Even a tiny gap of 0.5 mm can lead to arcing or heat under load. That’s why manufacturers focus heavily on connection requirements. The missing two sense wires aren’t the issue—they only inform the GPU about power limits. What often causes instability is micro-resistance from connectors that appear seated but aren’t fully locked. Loki’s early batches showed inconsistent quality in crimps; if one sense or power pin shows higher resistance, voltage spikes from 4090/4080 will quickly expose it. If your Corsair remains stable with OCCT and gaming while Loki fails, the problem likely lies in the cable or connector. For extra confidence, monitor GPU telemetry via HWInfo or GPU-Z during stress tests. A sudden power cap drop at a crash indicates a sensing line issue. A complete black screen suggests power loss. In short, optional sense pins aren’t the main concern—quality and proper seating matter more. If Corsair performs well under OCCT plus gaming and Loki keeps failing, you can be confident it’s the cable or connector that needs fixing.
I've tried both new versions and everything is working fine. Just mention that the Asus Loki comes with an extra RGB cable for compatibility, otherwise they're very similar. (GPU cables have two more or fewer) It's now running smoothly with EXPO II and Resize-bar. The only minor hiccup is waking up from sleep mode to shut down again after about 10 seconds. Probably not related to the latest updates. You're very helpful—maybe share some tips on choosing a PSU? My thought was Loki delivers top performance without needing anything below platinum plus, while the Corsair one is at platinum but gold plus elsewhere. I doubt the small energy boost from Asus matters much since my setup isn't that power-hungry (5070ti and 9800x3d, maybe around 500-600W). I'll let you know after testing the old model again with the same monitor.
If the Corsair performed well under stress, I’d hold onto it for the long haul, minimizing future risks—even with Loki’s impressive specs. Your hardware won’t reach 1000W, so stability and quality control become more important than efficiency ratings. Looking forward to seeing the final results.
I revisited and adjusted the settings, checking the latest update. After testing the new Loki for an extended period, I noticed no problems—no crashes at all. To confirm, I switched to the previous GPU model, the MSI 5070 TI Shadow 3x OC, to see if it performed better. Overall, I’m pleased with the MSI GPU’s performance and might think about getting more parts or a new laptop in the future. The Asus TUF GAMING OC GeForce RTX 5070Ti 16gb still works fine; it has more plastic and slightly smaller fans, but nothing too bad. I was expecting a crash with everything set up, but so far it hasn’t happened.
Since I don’t have much time to carry these devices, I’ll replace the CPU and compare which one runs better. I’m leaning toward keeping the Asus unit, as it’s reliable based on my experience, even though it’s pricier and has fewer ports than the MSI option. The new MSI model is more expensive—about $100 more—and offers three slots compared to MSI’s two.
Regarding the power supply, I hoped for a clear outcome, but I noticed some cable issues. The old ones were messy and not well routed; one was twisted or bent. After cleaning carefully with proper tools, I checked everything again, but it might still have affected the PSU. The Loki cables are shorter than the Corsair ones (I’ve used Corsair before), and they look better inside. The RMe 2025 model seems to handle heat better overall.
The fans seem plasticier in the Loki setup, while the MSI ones appear more robust. I’m curious about the PSU—maybe the cables were misplaced? I also noticed the fan design is sleek, but the components themselves are cheaper and covered with extra plastic, which could make the PSU hotter than the Loki. The grill looks good, though.
RAM and the case have always worked perfectly, even with small tweaks. I added a USB-C cable directly to the board for better connectivity, though it sometimes disconnects. The box arrived damaged, so it might be slightly affected.
I also considered whether the voltage fluctuations at EXPO could have caused issues, but I’m not too sure. For the final benchmarks, I plan to use CPUZ and FurMark (versions 2.16.0 and 2.9.0.0) and try swapping the PSU. I’ll overclock the CPU slightly at the end, though results so far are modest.
Overall, everything seems stable now, but I’m still curious about the CPU’s role in any potential problems.
Great work with the testing so far—it really helps pinpoint the issue. Since the Loki stayed stable, PSU seems less likely to be the cause. I see you’re using the Crucial kit at 6000 CL40 with EXPO; those parameters can sometimes be tricky. It might be a good idea to disable EXPO or adjust the timing slightly to eliminate that possibility. Also, testing the 9800X3D variant could help determine if the problem lies on the CPU side or in memory.
Just to clarify, I'm evaluating three PSUs. Two units of Loki and one Corsair. The one that worked without issues was the second Loki, while the first had cable problems. I think the fault likely came from those cables on the initial unit. During my last test, when I connected them, it seemed to establish a solid connection—possibly stronger than before. This explains why swapping the CPU caused crashes twice, but after swapping the PSU everything stabilized. The previous CPU tests (including two more 9800X3D units) all passed without problems. The RAM stayed consistent across all changes, and even with EXPO or JEDEC updates, nothing triggered crashes. Now the system runs fine with the old PSU, showing no signs of failure. This suggests the original issue was indeed related to the first PSU's poor cable connection. I'm leaning toward the new CPU being the better choice for performance, especially since it handles loads better and the multi-core stats are solid. I'll keep an eye on the GPU swap later to confirm stability. All current data points support moving forward with the updated setup.
It seems you've effectively identified the key points, especially the PSU cable details. The CPU and RAM appear to be in good shape, which means the final GPU change should validate everything. I'm hoping the system runs smoothly and you can confidently finalize your new setup without any unexpected issues.