The query asks whether a 3.3V rail being below the negative threshold could affect an NVMe SSD during gaming.
The query asks whether a 3.3V rail being below the negative threshold could affect an NVMe SSD during gaming.
Hi All
My 3.3 V rail reading shows 2.968V, can this influence the read/write functions of a NVMe SSD during gaming?
I also produce content where videos are around 40GB for a one-hour clip on YouTube. I use Davinci Resolve and don’t encounter any problems unless I play games that heavily read from the NVMe, such as Wukong, Indiana Jones, or SOTTR. Then I get an BSOD with Windows hardware uncorrectable error. Checking the logs points to a PCI root hub 3. Sometimes the game crashes and the SSD appears in the explorer but can’t open any folders because the path is invalid. Shutting down and restarting fixes the issue until it happens again.
The 12V reads as 11.952V and 5V drops to 4.95V.
Details: OS -win 10 latest, motherboard Gigabyte B660M, CPU i5 12400, RAM DDR4 16GB*2, one NVMe (256GB) for OS and another NVMe (1TB) for gaming. PSU is Corsair RM750x. I also have an HDD with movies, which works fine. My intuition suggests replacing the PSU – should I do that?
You need to verify how the voltage is being measured. The precision of the sensor matters. With a calibrated digital volt meter offering 4-digit accuracy, you'd trust the reading more.
It matches the BIOS reports and the gigabyte's S/W data as well. I'm considering testing the PSU with a multimeter. If your system contains this value, would you consider replacing the PSU?
This could be useful. You may want to press down the 24-way ATX PSU connector more firmly to enhance the connection with the motherboard.
https://www.techpowerup.com/forums/threa...ed.229795/
so after much discussion with others and using AI too !
😀
i chose to swap the PSU and so far the system hasn’t crashed. however, the 3.3 rail value remains below the expected level. this indicates the censor isn’t working properly at all. i plan to test the old PSU with a multimeter and check the rails. the outcome suggests the PSU lacks stability under varying loads. the games i was playing were wukong and SOTTR, so they clearly require different power levels. the PSU is still sufficient for rendering tasks. i’ll share my test results and conduct further checks by playing the games myself. hopefully this resolves the problem.
Sensors can serve as a reference, but they shouldn't be considered completely reliable. Using them will likely improve the accuracy of your measurements, especially if the meter was recently calibrated at a lab (as mentioned by @kanewolf). With a moderately reliable $3 AliExpress multimeter and calibration, I wouldn’t expect exceptional precision. I wouldn’t jump to conclusions unless I tested the PSU under various loads. Without continuous voltage monitoring during tests, it remains speculative about the root cause.
The discussion last night suggested a faulty connection with high resistance in the user's 24-way motherboard power connector. You might read 3.3 Volts at the PSU plug, but if the internal contacts are damaged, you could see only 2.9 Volts on the motherboard. The PSU’s voltage regulation isn’t necessarily bad—just the connection quality is an issue.
I don’t know which GPU you have, but if it’s from NVidia, are you using the Studio Driver or the Gaming Driver? I’m currently using Adobe, Topaz, and DaVinci, and I’m running the NVidia Studio Driver, which is known for better stability. AMD also offers comparable professional or gaming drivers for their graphics cards.
Given that your system runs smoothly in DaVinci, it’s unlikely the PSU instability is the main problem. The issue might lie elsewhere.
It's likely already mentioned, but you shouldn't rely solely on built-in sensors on the motherboard and operating system. You need to purchase a good digital multimeter, understand how to use it, and perform tests yourself. If any part of the power supply or AC/DC circuit deviates by more than 5%, it will definitely impact performance. This tolerance is typical for home and small business electronics, though it's tighter for high-end or military-grade devices.
A meter is also suitable for measuring on-load voltages. This is why I recommended conducting a series of tests at various loads to determine if the OP's PSU is defective. All the DC voltages at the output of an ATX PSU remain well below the SELV limit. For more details, refer to the specifications regarding maximum and minimum voltages for an ATX v2.2 PSU. Power Supply Voltage Tolerances (ATX v2.2) – PSU Tolerance Table – Voltage Rail, Minimum and Maximum Voltage values are provided.