Need help diagnosing random hard freezes after upgrading to an RTX 5070
PC Specs
New components (purchased in late 2025 / early 2026)
- CPU: AMD Ryzen 5 5500X3D (purchased Nov 2025)
- RAM: 32 GB (2x16 GB Kingston Fury Beast DDR4-3200 CL16, purchased Nov 2025)
- GPU: Palit GeForce RTX 5070 12 GB (purchased March 2026)
Older components
- Motherboard: Gigabyte B550M AORUS ELITE rev. 1.0 (purchased in 2021)
- PSU: GameMax GP650 650W 80 Plus Bronze (purchased in 2021)
- SSD: WD Green NVMe 1 TB (purchased in 2022)
- Backup GPU for testing: Zotac RTX 3060 12 GB (purchased in 2022)
- HDD: Old SATA HDD from 2012 (currently disconnected)
- Monitor: ASUS TUF 27" QHD 180 Hz (purchased in 2024)
When the issue started
My PC was infected by malware that stole all my passwords. I performed a clean Windows reinstall, and shortly after that I began experiencing random freezes.
Symptoms
Most of the time:
- The entire system freezes.
- Keyboard and mouse stop responding.
- Win + Ctrl + Shift + B does nothing.
- I have to hold the power button to shut down.
- Event Viewer only shows:
- Kernel-Power 41
- Sometimes volmgr (failed to create memory dump)
In one test using my old RTX 3060, the PC went to a black screen and rebooted automatically while still in the Battlefield 6 menu.
Workloads that trigger the issue
- Battlefield 6
- Hades II
- CapCut (editing and exporting videos)
Troubleshooting already performed
- Clean installation of Windows.
- BIOS updated/reflashed.
- Windows Memory Diagnostic: no errors.
- CrystalDiskInfo: SSD health normal.
- Temperatures are normal.
- OCCT tests completed successfully:
- CPU
- GPU
- RAM
- Power Test
- Disabled BitLocker.
- Removed a USB hub.
- Updated NVIDIA driver.
- Disabled NVIDIA Reflex.
- Forced the GPU slot to PCIe Gen4 (instead of Auto/Gen5).
- Disconnected an old 2012 SATA HDD.
Reports from other RTX 5070 users
I found multiple reports online from RTX 5070 owners experiencing similar random crashes, black screens, and hard freezes.
Some of the workarounds reported by other users include:
- Disabling NVIDIA Reflex.
- Forcing the PCIe slot to Gen4 instead of Auto/Gen5.
- Reinstalling NVIDIA drivers using DDU.
- Disabling overlays.
I have already applied some of these changes, but the freezes still occur intermittently.
Important discovery
I realized that my RTX 5070 was powered using one PCIe cable with a daisy-chain splitter, connected to the adapter included with the card.
From what I’ve read, the recommended setup is:
- Two separate PCIe cables, or
- A modern ATX 3.1 PSU with a native 12V-2x6 cable.
Current status
I temporarily switched back to my RTX 3060 while I wait for a new 850W PSU.
Current hypotheses
- GameMax GP650 PSU is unstable or insufficient.
- Using a single daisy-chain PCIe cable caused voltage drop under load.
- Defective RTX 5070.
- Motherboard issue.
- NVIDIA driver issue or RTX 50-series compatibility problem.
Questions
- Have you seen hard freezes caused by powering a GPU with a daisy-chain PCIe cable?
- Does passing all OCCT tests significantly reduce the likelihood of a PSU issue?
- Given that the problem also happened once with my RTX 3060, how likely is the PSU versus the motherboard?
- What additional tests would you run before requesting an RMA for the RTX 5070?
Any help would be greatly appreciated.