Diagnosing Freeze Issues Post-Installation (NVMe & CPU)
Diagnosing Freeze Issues Post-Installation (NVMe & CPU)
This situation is quite unusual. I recently added a new NVMe PCIe4 card and paired it with a Noctua n15 cooler. The system would suddenly freeze after some use. It was like the screen froze at one image, all sounds ceased, and only the reset or power button responded. After the freeze ended, I suspect I know how to initiate it, but I need more solid advice from experienced users. Could you describe your setup in more detail?
Initially, after installing the new components, I began partitioning the drive into three sections—planning to use one for general storage, another as a cache, and a third for games. I moved a game to the NVMe drive and started playing. About an hour later, the computer halted completely. At first, I assumed it was the CPU, since I bought it used, but I tried again. This time I checked the system stats and monitored the NVMe temperature. It reached 46°C, which isn’t extreme, but I’m concerned about the drive’s health.
I heard that Gen4 drives can overheat easily, and I’m using the included heatsink from the motherboard. The issue repeated, but no significant spike appeared in CPU temperatures. The NVMe stayed within normal limits. I moved the game back to my old SSD, which ran smoothly without freezing. Even overnight streaming didn’t trigger any problems.
My main worries are:
- Is the NVMe unit defective, possibly failing under load?
- Am I exceeding its thermal capacity with too many lanes?
- Could SATA connections be causing conflicts?
I found a manual chart of the connectors but can’t interpret it well. The drive is positioned directly beneath the Noctua fan, which is frustrating. I’m also out of thermal paste and don’t have an alcohol wipe handy. Returning it might cost more if it turns out to be faulty.
I’m considering transferring files to the new NVMe to free up SATA slots, but if it’s indeed bad, moving data could cause problems with linked project files. I’d appreciate any insights before making a decision.
I'm asking this because it's unclear what you're referring to. Could you clarify your question?
The cache drive helps with video editing by speeding up rendering and caching. It’s set up this way mainly for personal convenience. Partitioning multiple drives was done just to keep things organized in my mind. It can be confusing for others, but I prefer having everything neatly separated. If it caused issues, I guess I had to make it work.
Before installing a new CPU, verify your BIOS is current. If it was already updated, ensure you completely reset the CMOS. A 750W power rating might be insufficient and could become problematic later. Using SSD partitioning is generally ineffective and complicates controller operations, introducing more instability and potentially reducing performance. It may also prevent TRIM from functioning properly.
This introduces the lifespan and efficiency of solid-state storage systems. Contemporary SSDs function adequately without this feature, though performance may gradually decline over time due to the retention of unused data. It complements the controller’s overprovisioning strategy, ensuring balanced usage and reducing failure risks. While not critical, it adds value—especially since it’s included in Windows 7 and later without additional installation. Details about PCI-e connections require deeper analysis. Generally, systems operate smoothly or encounter issues immediately. The power supply unit’s age matters, along with the presence of a CPU and GPU. Typically, these components draw about 650W, though the GPU can cause spikes, nearing the limit. The NVMe SSD contributes an extra 10–15W, which may lead to instability if not monitored closely.
The PSU has reached its eleventh year, making it quite old. I previously discussed this on the forum a few months ago while thinking about upgrading to a new GPU. At that time, the advice was generally positive. However, before considering a new CPU and NVMe drive, I wanted to revisit the subject. How can I ensure my power supply remains suitable for future GPUs? Since they continue to grow in power requirements, I plan to switch GPUs again in about five to six years. The 40 series models are already pushing around 1000W. Regarding the components, besides the drives, there should be a USB expansion port—currently four are being used there.
I would measure the power draw with a wall meter. For long-term use, I usually overspec by about half, planning to keep the PSU for years. After ten years you might not have enough capacity left. In the future I’d likely need an 850W unit for such a setup, but if you want to stay ahead, 1000W or higher could become necessary. Unfortunately, prices have risen sharply for PSUs last year, making those options costly. Hopefully, next-generation graphics cards will become more efficient as well (unlikely).
I'll check if we have a wall meter at work. If not, it might take a while before I can inspect anything. We don't usually find many items for sale nearby. It seems the most likely issue is a lack of power—insufficient power could make the machine freeze. A possible fix right now is to disconnect some drives to lower power usage. Oh, and by the way, what does resetting CMOS do?
I was using Outer Vision's calculator for a rough idea. I connected more devices than usual just in case. The output I got looks normal and not overly concerning. It seems the advice about higher PSU wattage might be correct—excess power could affect performance. You might need another 750 or even 850 watts.