Occasional freezing and abrupt FPS reductions during gameplay on AMD Ryzen 7 with Nvidia RTX and Windows 10 Pro
Occasional freezing and abrupt FPS reductions during gameplay on AMD Ryzen 7 with Nvidia RTX and Windows 10 Pro
Hello, Welcome! Please note I might need to include more details in my next update. This concerns an unusual issue I face with my custom-built desktop: during gameplay, I experience a rapid drop in frames per second and the entire game freezes briefly—usually just a second or two. Most of these freezes last less than a second. I play Counter-Strike: Global Offensive. My build specs are: AMD Ryzen 7 3800X, MSI MPG X570 motherboard, Corsair iCUE RGB PRO XT with standard fans swapped for Noctua NF-F12 iPPC-2000, G.Skill Trident Z RGB memory (3200MHz), Nvidia RTX 2070 Super, LG Ultragear 27GL850 monitor, and Dell High Performance 27" storage. The operating system runs on a dedicated M.2 NVMe (Samsung 970 Evo) and the game is installed on another M.2 NVMe (Samsung 970 Evo Plus). My keyboard is Razer Huntsman Elite, mouse is Corsair M65 Elite, power supply is Razer Huntsman Elite, and the PSU is NZXT 750W. All drivers, including BIOS firmware, are up to date.
Previously, I used a Corsair CX750M power supply which lacked the extra 12V power, so I upgraded to a NZXT PSU. My older memory (Corsair Vengeance DDR4 32GB 2400MHz) was slow and had a higher CL value; I replaced it with G.Skill Trident for better performance. The new RAM is listed as compatible with my motherboard.
The problem continues. To observe the lag, watch this YouTube video. An interesting observation: when I restart the machine, the lag disappears temporarily but returns once the system stays on. Additionally, the lag isn’t confined to games—it appears during general tasks like browsing Chrome or typing. This is really frustrating and makes me very upset. I’m convinced there’s something missing—possibly hardware-related that I missed while assembling the system.
I don’t see similar issues on my Windows 10 gaming laptop, which helps point the problem back to my desktop build. Any suggestions would be greatly appreciated. Thanks for your time!
Curious about temperature effects, can you check probe readings? Is the device simply toggling or does its behavior change over time with machine use? Does it only activate when a specific condition is met, making it a persistent issue regardless of whether that trigger is present? How demanding is this operation? Will it diminish with extended use on simpler tasks?
Thanks for reaching out Bombastinator. My earlier concern about heat was resolved by switching to a Corsair iCUE RGB PRO XT cooling unit and moving to a Full ATX tower. Thermal readings stayed within normal range, and I’ve been monitoring them with CPUID Hardware Monitor Pro. The system never exceeded 72°C, even during extended use. For everyday tasks like browsing or video calls, the performance remains stable. From a communication standpoint, Discord works fine. I’ve disabled or closed background apps after reboots, but the issue persists. Performance tests using Perfmon and resource monitors didn’t pinpoint the cause. Any insights?
No thermal issues were found by the probes. Often, devices needing higher temperatures than displayed can cause problems eventually, since heat tends to spread and reach a probe. The gradual increase over time matches a heat-related concern, though memory leaks could also be responsible. A leak would show as rising memory usage, becoming problematic when swapping occurs because performance drops significantly. Slowdowns aren’t due to heavy heating, but they’re still a warning sign. It doesn’t rule out heat entirely, but suggests other factors should be checked. This issue has been present for over a year—did it begin with the initial setup? Were there changes to the power supply or RAM?
Apologies for the delayed reply. I hope you had a pleasant break over the holidays. So far, no thermal issues have arisen. Just to confirm, I refreshed the heatsink compound and reapplied it. The CPU temperature remains within its previous range. To further verify, I upgraded the cabinet from mid-tower to full-tower, ensuring ample airflow. The fans operate in a pull-push configuration: three pull fans draw cool air in, while one fan at the back expels it. The AIO radiator is set for push mode with a four-fan arrangement, creating an active airflow system. Regarding your second point, I own two units. The first model was built three years ago and still faced similar problems. After about six months, I replaced all components and built a second machine in anticipation of avoiding sudden freezes—though it continues to experience them. RAM was recently swapped from Corsair Vengeance DDR4 32GB 2400MHz (CMK16GX4M2A2400C16) to G.Skill Trident Z RGB Series 3200Mhz (Memory Model F4-3200C14D-16GTZR). This change was made because G.Skill offered a lower CL and higher frequency. XMP was enabled in the BIOS to allow the memory to reach 3200Mhz. The power supply was also replaced around the same time as the memory upgrade (during the full-tower installation from Corsair CX750M to NZXT 750W). My guess is that the PSU might be slightly under-specified, even though the guide suggests 750W should suffice.
It might help to look into ping values. If these are multiplayer titles, packet loss could reach the server or cause lag beyond your control. Multiplayer sessions run on specific platforms. If not, possible causes include a waiting state or system issues on your end. Heat might be involved, but other factors like resource contention or software conflicts could also play a role. Sudden memory usage spikes or excessive CPU activity may indicate problems. Correlating these events with timing data could offer clues. Lag patterns between certain times and cores can help identify the issue. Monitoring gameplay with a stopwatch might be useful, though it’s not ideal.
Hello everyone, I'm reaching out late with this update. I originally had tickets with both Nvidia and AMD, but nothing resolved. In the meantime, I adjusted some settings on my desktop. Here are the key changes in bold. After switching to a new board and NVMe drives, I reinstalled the OS, yet the issue persists. First, I filed a ticket with Nvidia, then after a month of back-and-forth troubleshooting, they replaced my graphics card with a refurbished RTX 2070 Super Founders Edition. The problem disappeared temporarily, but it returned shortly after. Later, I contacted AMD and after 65 days of investigation, they mentioned an existing fTPM issue with their processors that could cause intermittent slowdowns. The article reference is PA-410. I still doubt this matches my situation since I always used a hardware TPM (dTPM) or TPM 1.2. I keep experiencing problems and suspect there might be a configuration error or something wrong with the software side. I’d appreciate trying to pinpoint the exact cause and fix it. My frustration is high—I’m seriously considering giving up my desktop and buying a different machine, like an Intel Pentium 3 based one. Windows 10 won’t work, and my apps don’t run on Windows ME or Windows 98. If you’re reading this, I hope this becomes an interesting case study for all of you.
It isn’t a hardware issue, nor is it a problem with a specific piece of hardware you haven’t altered. You mentioned reinstalling everything. Was this a clean installation? What remains unchanged? The location of the computer, the power source, and the drives are identical. Are the storage devices, memory, and processor the same? (I doubt any of these are likely responsible) Check drivers, system configuration, and installed applications first. Have you checked the Task Manager to see if other apps consuming high bandwidth or frequent cycles were active simultaneously?